 |
View Old Forum Thread
Old Forum Index » Football Beta Testing » Football Beta Test Discussion » KLKRTR's Guide to Balancing the Sim: Part I
KLKRTR
*I may not finish each part in a single night (I'm starting this one at a quarter to one in the morning for example), but I will post what I have at bedtime and edit the post to add anything further I have the next day.
The data is on Post #5 in a downloadable Microsoft Excel 2010 spreadsheet, and I have also attached a link to the Google Docs version [url=https://spreadsheets.google.com/spreadsheet/ccc?key=0AjY-aczVe8C3dFNxdXc4dGZRNC0wQ3doN01taHd4Wmc&hl=en_US]here[/url].
[size=15]Part I: The Offense[/size]
[size=5]Table of Contents[/size]
[size=5]1. Passing[/size]
[size=4]? Beta League Stats
? Randy Moss League Stats
? NFL Stats
? What is balanced?[/size]
[size=3]- Completion Percentage
- Yards/Attempt
- Yards/Completion
- Passing 1st Down Conversions[/size]
[size=4]? What is not balanced?[/size]
[size=3]- Passing Attempts
- Interceptions
- Sacks
- TDs
- Passer Rating[/size]
[size=4]? Solutions[/size]
[size=5]2. Rushing[/size]
[size=4]? Beta League Stats
? Randy Moss League Stats
? NFL Stats
? What is balanced?[/size]
[size=3]- Yards/Carry
- Percent of Fumbles Lost[/size]
[size=4]? What is not balanced?[/size]
[size=3]- Rushing Attempts
- Fumbles
- Touchdowns
- Rushing 1st Down Conversions[/size]
[size=4]? Solutions[/size]
[size=5]3. Receiving[/size]
[size=4]? Beta League Stats
? Randy Moss League Stats
? NFL Stats
? What is balanced?[/size]
[size=3]- Yards/Reception[/size]
[size=4]? What is not balanced?[/size]
[size=3]- Yards After Catch Average
- Yards After Catch-to-Total Yards Percentage
- Touchdowns[/size]
[size=4]? Solutions
? Breakdown of Receiving Yardage by Receiver for 2010 NFL Season[/size]
[size=5]4. General Offensive Statistics[/size]
[size=4]? Beta League Stats
? Randy Moss League Stats
? NFL Stats
? What is balanced?[/size]
[size=3]- Yards/Play
- 3rd Down Attempts[/size]
[size=4]? What is not balanced?[/size]
[size=3]- Total Offensive Plays
- Passing-to-Rushing Ratio
- Total 1st Downs
- Penalty 1st Downs
- 3rd Down Conversions
- 3rd Down Conversion Percentage
- 4th Down Attempts
- 4th Down Conversions
- 4th Down Conversion Percentage[/size]
[size=4]? Solutions[/size]
[size=5]5. Concluding Remarks[/size]
-----------------------------------------------------------------------------------------------------------------------------------------------------
[size=5]Section 1: Passing[/size]
Ok, so some would say that the NFL is a passing league right now. That's probably true, judging by the stats, but if passing is done so much (and has been statistically proven to have more upside than rushing between the 20s), then why do we not have a balanced passing attack in the SD leagues? Part of that is certainly the limited play-calling available to team owners as of right now, but that is also the case for rushing, so that's not all of it. SD quarterbacks in this study have an average passer rating of 87.2--five and a half points higher than the average NFL QB over the last two seasons. Well, there are two balances that should be pretty easy to implement into the sim that would help with the above-average QB rating (which is based off of completion percentage, yards/attempt, touchdowns, and interceptions). The first balance has already been put in place by Chris, so we'll get to see how 150% of the previous amount of interceptions affects the game there. The other major stat that is way off in passing is the amount of times quarterbacks get sacked. In the NFL, QBs got sacked 6.5% of all pass attempts. In our leagues, there were sacks on only 4.3% of the plays (a difference of almost 1 sack per team per game). The two remaining issues, passing attempts (about 4 too few attempts per game per team) and passing TDs (about 5.5 TDs too few per season per team), are a bit more complicated to fix, and would require adjustments across the board. We'll examine this more in Section 5 (Concluding Remarks), but keep this final stat in mind: The number of TDs per season would be expected to be lower because of the lower number of passing attempts, but what causes NFL teams to score TDs on 4.3% of their pass attempts (to SD's 3.8%)?
[size=5]Section 2: Rushing[/size]
Rushing is pretty high-powered in SD. I think we can all agree with that. But just how much more powerful is the average SD rushing attack compared with an NFL rushing attack? Well, SD offenses score rushing TDs about 23% more often on rushing attempts than NFL offenses (3.7% to 3.0%). That's the exact opposite of what we saw in the passing game, so not really a surprise, but I'm sure there is something we can do to eliminate all or part of that discrepancy. I wanted to check for tackle for loss (TFL) stats, but I could not find them for the NFL. If anyone knows where to find those, I'd love to see a comparison there. Now, there are obviously more carries per game in SD (not quite 2 more per game), but the good news is that the yards per carry (4.2) is the same for both leagues. Those two stats combined gives SD teams about 114 rush yards more per season. Now, here's what surprised me: SD's rushing attack actually needs to get better. You heard me right. That's not a typo. SD teams are averaging 11 fumbles (lost and recovered combined) per season, but NFL teams are averaging only 7.8. In case you're trying to do the math, that's 33% more often in SD. That leads to an average of about 2 more turnovers for SD teams (however, the percentage of fumbles that are not recovered is in line with live stats). Again, this should be a pretty easy fix. Now if only we could figure out why SD players run more than pass (despite averaging almost 3 yards more per pass than rush). Like I said above, we'll look at this more in Section 5.
[size=5]Section 3: Receiving[/size]
Ok, so receiving is covered, for the most part, by what was already said in the passing guide, so I'll try to avoid covering much of that, but at the same time, there's something here. In terms of yards/team, we've already stated that NFL teams average about 350 receiving yards more than the average SD team. But what are the parts of a reception?
1) The Reception. Hard to judge how this compares without drop stats for the leagues.
2) The Receiving Yards. Too few in SD, but the average yards/reception is about right (11.0 in SD; 10.8 in NFL). Therefore, it would seem the only discrepancy is the number of receptions, if it weren't for part 3.
3) The Yards After Catch (YAC). In the NFL, YAC accounts for ~49% of total receiving yardage (meaning half the passing yardage comes on the ground). In other words, for the average 10.8 yard reception, 5.3 of those yards are YAC. Now, here's where the problem is: SD teams average about 3.4 YAC/reception--equivalent to ~31% of the total passing yardage. BIG difference there (almost 20% difference).
In my opinion, if we were to increase the amount of YAC/reception (and lowered the yards in the air accordingly), then you would see more passing TDs. I know, for me at least, my red zone offense is mostly rushing (and therefore leads to a larger number of rushing TDs), but I don't get too many passing TDs outside of the red zone now anymore either. With increased YAC, teams would be more likely to score from further out (which would improve upon the passing TD-to-rushing TD ratio. With teams scoring more in passing, I have a feeling there would be at least a slight shift back towards passing more as well.
Receiving Yards Broken Down by Receiver Type
#1 WR AVG Game: 7.8 Passes/G; 62.3 Yd/G (8.0 ypa)
#2 WR AVG Game: 6.2 Passes/G; 48.4 Yd/G (7.8 ypa)
Other WR AVG Game: 6.0 Passes/G; 44.5 Yd/G (7.4 ypa)
TE AVG Game: 7.0 Passes/G; 49.7 Yd/G (7.1 ypa)
RB AVG Game: 6.6 Passes/G; 38.6 Yd/G (5.8 ypa)
I want to try and do something similar for SD, but that takes time...
Looking at just one team though (randomly selected from the middle of the pack in the RMFL), we get the following:
#1 WR: 11.2 Passes/G; 79.5 Yd/G (7.1 ypa)
#2 WR: 9.2 Passes/G; 69.3 Yd/G (7.5 ypa)
Other WR: 3.7 Passes/G; 15.3 Yd/G (4.1 ypa)
TE: 3.6 Passes/G; 16.9 Yd/G (4.7 ypa)
RB: 3.0 Passes/G; 13.1 Yd/G (4.4 ypa)
It's not an average, but it gives us an idea of what direction we need to go in. Basically we need to pass to the #1 and #2 WRs so much. You have to realize that "Other WR" is actually 3 guys, "TE" is 2 guys, and "RB" is typically 3 guys (1 starter, 1 backup, and 1 FB).
[size=5]Section 4: General Offensive Statistics[/size]
Ok, just a bunch of seemingly random statistics (mainly ones I could find and/or calculate and compare), but here goes. Like has been stated multiple times by now, the passing-to-rushing ratio is out of whack, but we'll see where that goes with the screen percentage increase and a proposed change in YAC. NFL teams are averaging about 43 plays per season more (almost 4 a game), but that's possibly attributed to more rushing (I don't know)?
Now, lets look at downs. NFL teams average from 53 to 54 more first downs per season than SD teams (4 first downs a game is pretty significant, no?). Now, all of those extra first downs, where do they come from? SD teams appear to average about 11 fewer 1st downs by penalty per season. That would be a big one. As for which penalties, I'm assuming encroachment on shorter downs and a couple more pass interference calls, but that's a complete guess. I just don't think it's from defensive holding. Probably personal fouls are up there as well. Where else do 53 first downs come from? Well, it doesn't appear to be from 3rd down conversions, as SD teams actually convert about 5% more often on a similar number of 3rd downs. Now here's the other thing: 4th downs. It's not a huge factor in terms of 53 extra first downs, but it's a big deal in games usually (or at least has the potential to). Normally, convention tells us that the computer player will be more of a risk-taker than an actual coach with his job on the line, but here are the actual facts:
NFL teams attempt an average of more than 3 extra 4th downs. They average over 4 more conversions. Part of this is play-calling, but that's actually a big thing. NFL teams convert half of their 4th down attempts; SD teams only 30%. Therefore, it makes sense to not go for them as often. The reason the percentages are so different has to be the way 4th downs are decided upon. The current system is, in my opinion, a poor way to choose which 4th downs to go for. I know I'd much rather go for that 4th & inches on my own 35 than the 4th & 18 on the opponent's 35. My own idea would be to offer an option to go for it on 4th and short (define short) within a certain distance (define) from the goalline. I think that would really improve the game, and we'd see fewer drives stall out because the team came up because of coming up an inch short at the 50.
[size=5]Section 5: Concluding Remarks[/size]
Well, it took me a long time to put this together, and while most of the information is worthless to us, I think that there are a few keys in there that we hadn't looked at and/or considered to be problems. Like I said at the beginning, I want to do a similar post for both defense and special teams, so that will be forthcoming at some point, but once we can get the multi-season leagues going, we can incorporate that information into this as well, and I also think that Chris might then have some time to focus on the things like Sack %, Interception %, fumble %, YAC %, penalty 1st downs, and 4th down conversions. After that, maybe we could look at what can be done about balancing out the passing?
As always, I love to see discussion and comments! I try to incorporate anything you guys talk about into posts like these.
Also, Chris, for the next Beta season, how hard would it be to track receiver drops? Or blocking stats?
[Edited on 7-5-2011 by KLKRTR]
KLKRTR's Guide to Balancing the Sim: Part I
June 30, 2011 at 05:50AM View BBCode
Ok, first off, not trying to be cocky and say I have all the answers. I'm just going to try and figure out what are all of the problems, and where applicable, I will put any suggestions I have for fixing a specific problem. This will be my biggest project to date for a non-school related topic, so it could take me a while. I'm breaking down this "guide" into parts, starting tonight with Part I, which will deal solely with the offense. Next, I will try to do the same thing for the defense. After that, I will do special teams. Finally, the last section will be for miscellaneous items (i.e. penalties, time outs, etc.). This guide is not for non-balancing issues, as that is not the priority at the moment. Any comments and/or criticisms are not only welcomed, but encouraged--as always! Any and all data sets will exclude data from the Arkansas Abusers, as that data is worthless to us now pretty much except to see how a tanking team performs. "All-Time Beta League data" simply means data from the last two seasons only. Keep in mind that any stats compiled in two seasons in SD is normally equivalent to the same number of games in one NFL season; however, I eliminated Arkansas, so only averages are equivalent. Now, please (try to) enjoy!*I may not finish each part in a single night (I'm starting this one at a quarter to one in the morning for example), but I will post what I have at bedtime and edit the post to add anything further I have the next day.
The data is on Post #5 in a downloadable Microsoft Excel 2010 spreadsheet, and I have also attached a link to the Google Docs version [url=https://spreadsheets.google.com/spreadsheet/ccc?key=0AjY-aczVe8C3dFNxdXc4dGZRNC0wQ3doN01taHd4Wmc&hl=en_US]here[/url].
[size=15]Part I: The Offense[/size]
[size=5]Table of Contents[/size]
[size=5]1. Passing[/size]
[size=4]? Beta League Stats
? Randy Moss League Stats
? NFL Stats
? What is balanced?[/size]
[size=3]- Completion Percentage
- Yards/Attempt
- Yards/Completion
- Passing 1st Down Conversions[/size]
[size=4]? What is not balanced?[/size]
[size=3]- Passing Attempts
- Interceptions
- Sacks
- TDs
- Passer Rating[/size]
[size=4]? Solutions[/size]
[size=5]2. Rushing[/size]
[size=4]? Beta League Stats
? Randy Moss League Stats
? NFL Stats
? What is balanced?[/size]
[size=3]- Yards/Carry
- Percent of Fumbles Lost[/size]
[size=4]? What is not balanced?[/size]
[size=3]- Rushing Attempts
- Fumbles
- Touchdowns
- Rushing 1st Down Conversions[/size]
[size=4]? Solutions[/size]
[size=5]3. Receiving[/size]
[size=4]? Beta League Stats
? Randy Moss League Stats
? NFL Stats
? What is balanced?[/size]
[size=3]- Yards/Reception[/size]
[size=4]? What is not balanced?[/size]
[size=3]- Yards After Catch Average
- Yards After Catch-to-Total Yards Percentage
- Touchdowns[/size]
[size=4]? Solutions
? Breakdown of Receiving Yardage by Receiver for 2010 NFL Season[/size]
[size=5]4. General Offensive Statistics[/size]
[size=4]? Beta League Stats
? Randy Moss League Stats
? NFL Stats
? What is balanced?[/size]
[size=3]- Yards/Play
- 3rd Down Attempts[/size]
[size=4]? What is not balanced?[/size]
[size=3]- Total Offensive Plays
- Passing-to-Rushing Ratio
- Total 1st Downs
- Penalty 1st Downs
- 3rd Down Conversions
- 3rd Down Conversion Percentage
- 4th Down Attempts
- 4th Down Conversions
- 4th Down Conversion Percentage[/size]
[size=4]? Solutions[/size]
[size=5]5. Concluding Remarks[/size]
-----------------------------------------------------------------------------------------------------------------------------------------------------
[size=5]Section 1: Passing[/size]
Ok, so some would say that the NFL is a passing league right now. That's probably true, judging by the stats, but if passing is done so much (and has been statistically proven to have more upside than rushing between the 20s), then why do we not have a balanced passing attack in the SD leagues? Part of that is certainly the limited play-calling available to team owners as of right now, but that is also the case for rushing, so that's not all of it. SD quarterbacks in this study have an average passer rating of 87.2--five and a half points higher than the average NFL QB over the last two seasons. Well, there are two balances that should be pretty easy to implement into the sim that would help with the above-average QB rating (which is based off of completion percentage, yards/attempt, touchdowns, and interceptions). The first balance has already been put in place by Chris, so we'll get to see how 150% of the previous amount of interceptions affects the game there. The other major stat that is way off in passing is the amount of times quarterbacks get sacked. In the NFL, QBs got sacked 6.5% of all pass attempts. In our leagues, there were sacks on only 4.3% of the plays (a difference of almost 1 sack per team per game). The two remaining issues, passing attempts (about 4 too few attempts per game per team) and passing TDs (about 5.5 TDs too few per season per team), are a bit more complicated to fix, and would require adjustments across the board. We'll examine this more in Section 5 (Concluding Remarks), but keep this final stat in mind: The number of TDs per season would be expected to be lower because of the lower number of passing attempts, but what causes NFL teams to score TDs on 4.3% of their pass attempts (to SD's 3.8%)?
[size=5]Section 2: Rushing[/size]
Rushing is pretty high-powered in SD. I think we can all agree with that. But just how much more powerful is the average SD rushing attack compared with an NFL rushing attack? Well, SD offenses score rushing TDs about 23% more often on rushing attempts than NFL offenses (3.7% to 3.0%). That's the exact opposite of what we saw in the passing game, so not really a surprise, but I'm sure there is something we can do to eliminate all or part of that discrepancy. I wanted to check for tackle for loss (TFL) stats, but I could not find them for the NFL. If anyone knows where to find those, I'd love to see a comparison there. Now, there are obviously more carries per game in SD (not quite 2 more per game), but the good news is that the yards per carry (4.2) is the same for both leagues. Those two stats combined gives SD teams about 114 rush yards more per season. Now, here's what surprised me: SD's rushing attack actually needs to get better. You heard me right. That's not a typo. SD teams are averaging 11 fumbles (lost and recovered combined) per season, but NFL teams are averaging only 7.8. In case you're trying to do the math, that's 33% more often in SD. That leads to an average of about 2 more turnovers for SD teams (however, the percentage of fumbles that are not recovered is in line with live stats). Again, this should be a pretty easy fix. Now if only we could figure out why SD players run more than pass (despite averaging almost 3 yards more per pass than rush). Like I said above, we'll look at this more in Section 5.
[size=5]Section 3: Receiving[/size]
Ok, so receiving is covered, for the most part, by what was already said in the passing guide, so I'll try to avoid covering much of that, but at the same time, there's something here. In terms of yards/team, we've already stated that NFL teams average about 350 receiving yards more than the average SD team. But what are the parts of a reception?
1) The Reception. Hard to judge how this compares without drop stats for the leagues.
2) The Receiving Yards. Too few in SD, but the average yards/reception is about right (11.0 in SD; 10.8 in NFL). Therefore, it would seem the only discrepancy is the number of receptions, if it weren't for part 3.
3) The Yards After Catch (YAC). In the NFL, YAC accounts for ~49% of total receiving yardage (meaning half the passing yardage comes on the ground). In other words, for the average 10.8 yard reception, 5.3 of those yards are YAC. Now, here's where the problem is: SD teams average about 3.4 YAC/reception--equivalent to ~31% of the total passing yardage. BIG difference there (almost 20% difference).
In my opinion, if we were to increase the amount of YAC/reception (and lowered the yards in the air accordingly), then you would see more passing TDs. I know, for me at least, my red zone offense is mostly rushing (and therefore leads to a larger number of rushing TDs), but I don't get too many passing TDs outside of the red zone now anymore either. With increased YAC, teams would be more likely to score from further out (which would improve upon the passing TD-to-rushing TD ratio. With teams scoring more in passing, I have a feeling there would be at least a slight shift back towards passing more as well.
Receiving Yards Broken Down by Receiver Type
#1 WR AVG Game: 7.8 Passes/G; 62.3 Yd/G (8.0 ypa)
#2 WR AVG Game: 6.2 Passes/G; 48.4 Yd/G (7.8 ypa)
Other WR AVG Game: 6.0 Passes/G; 44.5 Yd/G (7.4 ypa)
TE AVG Game: 7.0 Passes/G; 49.7 Yd/G (7.1 ypa)
RB AVG Game: 6.6 Passes/G; 38.6 Yd/G (5.8 ypa)
I want to try and do something similar for SD, but that takes time...
Looking at just one team though (randomly selected from the middle of the pack in the RMFL), we get the following:
#1 WR: 11.2 Passes/G; 79.5 Yd/G (7.1 ypa)
#2 WR: 9.2 Passes/G; 69.3 Yd/G (7.5 ypa)
Other WR: 3.7 Passes/G; 15.3 Yd/G (4.1 ypa)
TE: 3.6 Passes/G; 16.9 Yd/G (4.7 ypa)
RB: 3.0 Passes/G; 13.1 Yd/G (4.4 ypa)
It's not an average, but it gives us an idea of what direction we need to go in. Basically we need to pass to the #1 and #2 WRs so much. You have to realize that "Other WR" is actually 3 guys, "TE" is 2 guys, and "RB" is typically 3 guys (1 starter, 1 backup, and 1 FB).
[size=5]Section 4: General Offensive Statistics[/size]
Ok, just a bunch of seemingly random statistics (mainly ones I could find and/or calculate and compare), but here goes. Like has been stated multiple times by now, the passing-to-rushing ratio is out of whack, but we'll see where that goes with the screen percentage increase and a proposed change in YAC. NFL teams are averaging about 43 plays per season more (almost 4 a game), but that's possibly attributed to more rushing (I don't know)?
Now, lets look at downs. NFL teams average from 53 to 54 more first downs per season than SD teams (4 first downs a game is pretty significant, no?). Now, all of those extra first downs, where do they come from? SD teams appear to average about 11 fewer 1st downs by penalty per season. That would be a big one. As for which penalties, I'm assuming encroachment on shorter downs and a couple more pass interference calls, but that's a complete guess. I just don't think it's from defensive holding. Probably personal fouls are up there as well. Where else do 53 first downs come from? Well, it doesn't appear to be from 3rd down conversions, as SD teams actually convert about 5% more often on a similar number of 3rd downs. Now here's the other thing: 4th downs. It's not a huge factor in terms of 53 extra first downs, but it's a big deal in games usually (or at least has the potential to). Normally, convention tells us that the computer player will be more of a risk-taker than an actual coach with his job on the line, but here are the actual facts:
NFL teams attempt an average of more than 3 extra 4th downs. They average over 4 more conversions. Part of this is play-calling, but that's actually a big thing. NFL teams convert half of their 4th down attempts; SD teams only 30%. Therefore, it makes sense to not go for them as often. The reason the percentages are so different has to be the way 4th downs are decided upon. The current system is, in my opinion, a poor way to choose which 4th downs to go for. I know I'd much rather go for that 4th & inches on my own 35 than the 4th & 18 on the opponent's 35. My own idea would be to offer an option to go for it on 4th and short (define short) within a certain distance (define) from the goalline. I think that would really improve the game, and we'd see fewer drives stall out because the team came up because of coming up an inch short at the 50.
[size=5]Section 5: Concluding Remarks[/size]
Well, it took me a long time to put this together, and while most of the information is worthless to us, I think that there are a few keys in there that we hadn't looked at and/or considered to be problems. Like I said at the beginning, I want to do a similar post for both defense and special teams, so that will be forthcoming at some point, but once we can get the multi-season leagues going, we can incorporate that information into this as well, and I also think that Chris might then have some time to focus on the things like Sack %, Interception %, fumble %, YAC %, penalty 1st downs, and 4th down conversions. After that, maybe we could look at what can be done about balancing out the passing?
As always, I love to see discussion and comments! I try to incorporate anything you guys talk about into posts like these.
Also, Chris, for the next Beta season, how hard would it be to track receiver drops? Or blocking stats?
[Edited on 7-5-2011 by KLKRTR]
Admin
The sim tends to suffer from "butterfly effect", as making an adjustment in one place affects everything else, and of course all the numbers are skewed by owner strategies, which in Beta are often extreme to stress the system. Because sim owners don't have the same pressures to play conservative, I think that some numbers may not match NFL numbers, especially on the special teams side.
That said, here are some comments on what you've posted:
- Passing attempts is purely based on strategies, and I suspect in the NFL this number varies wildly over the years as different strategic philosophies come and go.
- Completion rate will likely vary a couple of percent in either direction based on talent
- Yards/attempt and yards/completion are fairly close to reality
- Interceptions are of course way low, the new code should hopefully fix that
- I am surprised at the sack numbers, my gut thought was that sacks are too high.
- Any 1 or 2 seasons is a fairly small sample (including reality). Looking at the stats in the database for ALL teams on the live site gets me an average of 6.3 yards per attempt (vs. 6.6 for 2010) and 10.67 yards/completion (vs. 10.8), a total completion rate of 59.2 (vs. 60.8), and an interception rate of 2.3% (as opposed to 3.0%). The sitewide QB rating is 77.6, which is sort of at odds with Beta.
-Chris
June 30, 2011 at 06:52PM View BBCode
It's not cocky at all, I appreciate any help.The sim tends to suffer from "butterfly effect", as making an adjustment in one place affects everything else, and of course all the numbers are skewed by owner strategies, which in Beta are often extreme to stress the system. Because sim owners don't have the same pressures to play conservative, I think that some numbers may not match NFL numbers, especially on the special teams side.
That said, here are some comments on what you've posted:
- Passing attempts is purely based on strategies, and I suspect in the NFL this number varies wildly over the years as different strategic philosophies come and go.
- Completion rate will likely vary a couple of percent in either direction based on talent
- Yards/attempt and yards/completion are fairly close to reality
- Interceptions are of course way low, the new code should hopefully fix that
- I am surprised at the sack numbers, my gut thought was that sacks are too high.
- Any 1 or 2 seasons is a fairly small sample (including reality). Looking at the stats in the database for ALL teams on the live site gets me an average of 6.3 yards per attempt (vs. 6.6 for 2010) and 10.67 yards/completion (vs. 10.8), a total completion rate of 59.2 (vs. 60.8), and an interception rate of 2.3% (as opposed to 3.0%). The sitewide QB rating is 77.6, which is sort of at odds with Beta.
-Chris
redcped
For example, let's say there are 3 standout RB in the initial draft. Over the next 3 amateur drafts, let's say 6 other teams acquire standout RBs, too. And in the meantime, let's assume that normal improvement brought up the talents of 4 other RBs from the initial draft to a top level. So now, here we are a few seasons in, and almost every team has a superlative feature back.
Now do we know yet how talent inflation overall will balance out, so that essentially it becomes the norm for everyone to have top talent at many positions, and it all evens out? Maybe it will ... or maybe as we have seen in baseball sometimes leagues become very pitching dominant or hitter dominant based on draft quality and CP distribution. So we might see the rise of killer defenses, or unstoppable running games, or who knows what?
My point is that anyone who plays SD knows that 1950 is often an outlier season for the remainder of league history. Stats get posted in the first season that are never matched again. By the 1960s, things even out and you get a realistic sense of player performance in a mature league.
I'm not saying this thread lacks value at all. It's a great idea. I'm just pointing out that we might be far better served by accelerating into multi-season and seeing how talent develops over time and how that affects stats. We might not be seeing the sim at its mature stage.
June 30, 2011 at 09:24PM View BBCode
Something we need to keep in mind here is that multi-season play will affect a great deal of the statistics and we should be testing that before we ascribe too much to single seasons.For example, let's say there are 3 standout RB in the initial draft. Over the next 3 amateur drafts, let's say 6 other teams acquire standout RBs, too. And in the meantime, let's assume that normal improvement brought up the talents of 4 other RBs from the initial draft to a top level. So now, here we are a few seasons in, and almost every team has a superlative feature back.
Now do we know yet how talent inflation overall will balance out, so that essentially it becomes the norm for everyone to have top talent at many positions, and it all evens out? Maybe it will ... or maybe as we have seen in baseball sometimes leagues become very pitching dominant or hitter dominant based on draft quality and CP distribution. So we might see the rise of killer defenses, or unstoppable running games, or who knows what?
My point is that anyone who plays SD knows that 1950 is often an outlier season for the remainder of league history. Stats get posted in the first season that are never matched again. By the 1960s, things even out and you get a realistic sense of player performance in a mature league.
I'm not saying this thread lacks value at all. It's a great idea. I'm just pointing out that we might be far better served by accelerating into multi-season and seeing how talent develops over time and how that affects stats. We might not be seeing the sim at its mature stage.
KLKRTR
Also, I have asked about the possibility of a multi-season league becoming more balanced, and I'm sure it will, but at the same time, I don't think it would be drastic as, like you said, there would be more talent on each side of the ball. (Kicking stats would probably become better with time though)
Chris, I realize that, while different philosophies come and go, I will be using some of that data to make sure that there isn't a problem there. If there is, we'll be able to see it. If not, and it is simply due to strategies (this is partly why I am including the RMFL), then it can be crossed off the list pretty easily.
July 01, 2011 at 12:50AM View BBCode
I'm doing a lot of updating on the post tonight, but it probably won't be finished until the weekend because I decided to add data from 2 seasons in the Randy Moss League (to get a standard competitive/active league) and a second NFL season (to increase sample size).Also, I have asked about the possibility of a multi-season league becoming more balanced, and I'm sure it will, but at the same time, I don't think it would be drastic as, like you said, there would be more talent on each side of the ball. (Kicking stats would probably become better with time though)
Chris, I realize that, while different philosophies come and go, I will be using some of that data to make sure that there isn't a problem there. If there is, we'll be able to see it. If not, and it is simply due to strategies (this is partly why I am including the RMFL), then it can be crossed off the list pretty easily.
KLKRTR
Ok, it worked, so I'll have my stats on this Excel document (just doesn't look the same on Google Docs).
[Edited on 7-1-2011 by KLKRTR]
Ok, so before, I had only uploaded the spreadsheet with some of the passing stats. All passing, rushing, and receiving stats are up for all 3 leagues in the updated spreadsheet (attached in this post). Some of these numbers should open some eyes...unfortunately, I can barely keep mine open now at 20 til 2am, so I'm gonna head to bed and post the general offensive stats plus all of my comments (and there will be a lot of those) tomorrow evening. I never realized it would take me over 2 evenings to compile all this data, but apparently some stats are hard to find (or you have to add up individual players' stats--had to do that for 2009/2010 NFL receivers' YAC). If anyone knows a really detailed NFL stats site, I'd love to get the link.
Still to come for Part I:
? 2-pt conversions
? Offensive Points Scored/Game
? More Commentary
Fortunately, this spreadsheet should give us plenty to look at/discuss while I get that stuff up. I know there wasn't a lot to give feedback on after my first day (next to no stats up, but now we should have tables with sortable columns to compare the three different leagues (2 years each), the averages for each league, and a Beta+Randy Moss total.
[size=10]Passing, Rushing, and Receiving Excel Spreadsheet Here[/size]
(Just realized that I posted the wrong version last night...here's the real version:
[Edited on 7-2-2011 by KLKRTR]
[Edited on 7-2-2011 by KLKRTR]
[Edited on 7-2-2011 by KLKRTR]
July 01, 2011 at 01:55AM View BBCode
I just wanted to see if this would upload or not.Ok, it worked, so I'll have my stats on this Excel document (just doesn't look the same on Google Docs).
[Edited on 7-1-2011 by KLKRTR]
Ok, so before, I had only uploaded the spreadsheet with some of the passing stats. All passing, rushing, and receiving stats are up for all 3 leagues in the updated spreadsheet (attached in this post). Some of these numbers should open some eyes...unfortunately, I can barely keep mine open now at 20 til 2am, so I'm gonna head to bed and post the general offensive stats plus all of my comments (and there will be a lot of those) tomorrow evening. I never realized it would take me over 2 evenings to compile all this data, but apparently some stats are hard to find (or you have to add up individual players' stats--had to do that for 2009/2010 NFL receivers' YAC). If anyone knows a really detailed NFL stats site, I'd love to get the link.
Still to come for Part I:
? 2-pt conversions
? Offensive Points Scored/Game
? More Commentary
Fortunately, this spreadsheet should give us plenty to look at/discuss while I get that stuff up. I know there wasn't a lot to give feedback on after my first day (next to no stats up, but now we should have tables with sortable columns to compare the three different leagues (2 years each), the averages for each league, and a Beta+Randy Moss total.
[size=10]Passing, Rushing, and Receiving Excel Spreadsheet Here[/size]
(Just realized that I posted the wrong version last night...here's the real version:
[Edited on 7-2-2011 by KLKRTR]
[Edited on 7-2-2011 by KLKRTR]
[Edited on 7-2-2011 by KLKRTR]
KLKRTR has attached this file: SD Offense.xlsx (application/vnd.openxmlformats-officedocument.spreadsheetml.sheet)
RichNYC1
This is a very important point. Once we get a chance to develop players you will see better 1 on 1 match-ups, more even QB match-ups and the ability to draft to specific strategies, be it run, pass or defensive vs. offensive philosophy. I am a big believer in EXE, which should dramatically change (and I think help the defense).
As an example, look at tackling ability vs break tackling ability. These high Brkt guys have a limited ceiling while defenders (especially DB´s) have a lot of room for development. That?s going to change the passing game dramatically.
I think it will be very much like to SDBB, numbers in 1950´s vs. 20- 30 years later are always very different. You see OPS and HR´s drop as the pitching and defense gets better. Bottom line, I think we are on the right track but need to see multiple seasons play out
[Edited on 10/04/06 by RichNYC1]
July 01, 2011 at 03:33PM View BBCode
Originally posted by redcped
Something we need to keep in mind here is that multi-season play will affect a great deal of the statistics and we should be testing that before we ascribe too much to single seasons.
This is a very important point. Once we get a chance to develop players you will see better 1 on 1 match-ups, more even QB match-ups and the ability to draft to specific strategies, be it run, pass or defensive vs. offensive philosophy. I am a big believer in EXE, which should dramatically change (and I think help the defense).
As an example, look at tackling ability vs break tackling ability. These high Brkt guys have a limited ceiling while defenders (especially DB´s) have a lot of room for development. That?s going to change the passing game dramatically.
I think it will be very much like to SDBB, numbers in 1950´s vs. 20- 30 years later are always very different. You see OPS and HR´s drop as the pitching and defense gets better. Bottom line, I think we are on the right track but need to see multiple seasons play out
[Edited on 10/04/06 by RichNYC1]
Admin
--Chris
July 01, 2011 at 03:56PM View BBCode
It is worth noting that players here are generated differently than on the baseball side. All players, whether in the college draft or initial draft are generated with the same codfe. Those initial draft players who are older than college freshmen (i.e., all of them) then are accelerated through enough seasons of improves and declines to get them to their correct age. So theoretically there shouldn't be very much actual talent inflation. In reality therre will be, as directed development will give better results than the algorithms that are meant to simulate development, but it hopefully won't be a dramatic difference.--Chris
KLKRTR
Should give you guys something to read on your holiday weekend
Happy 4th!! Enjoy it guys
July 02, 2011 at 05:58AM View BBCode
The original post (and the Excel Sheet in post #5) have been updated.Should give you guys something to read on your holiday weekend
Happy 4th!! Enjoy it guys
KLKRTR
Found a few things that need to be looked at, and a couple of other things to discuss.
July 05, 2011 at 05:35AM View BBCode
Post is finished. Let me know what you guys think.Found a few things that need to be looked at, and a couple of other things to discuss.
RichNYC1
I would think after 5, 6, 10 seasons there will be a significant difference
July 05, 2011 at 04:02PM View BBCode
Originally posted by Admin
In reality therre will be, as directed development will give better results than the algorithms that are meant to simulate development, but it hopefully won't be a dramatic difference.
I would think after 5, 6, 10 seasons there will be a significant difference
KLKRTR
July 05, 2011 at 11:26PM View BBCode
Since there will be directed development for all players though, I don't see such a significant difference (obviously, on a per-team basis, there will be plenty, but I see it averaging itself out)
Fulla
4th down conv % is slightly lower than the NFL. My personal settings are more aggressive on 4th down than most NFL coaches.
[Edited on 7-6-2011 by Fulla]
July 06, 2011 at 03:57AM View BBCode
3rd down conv % is slightly higher than the NFL. I find it difficult to stop a team on 3rd & any distance. Both Beta & RMFL increased their 3rd down conv % in the second season.4th down conv % is slightly lower than the NFL. My personal settings are more aggressive on 4th down than most NFL coaches.
[Edited on 7-6-2011 by Fulla]
KLKRTR
You say your 4th down settings are more aggressive than NFL coaches, and I agree with that for "going for it" on your opponent's half, but how often do you go for it on 4th & inches/short on your own half (compared with NFL coaches)?
I can answer it for you: you can't go for it. That's a major flaw in my eyes.
July 06, 2011 at 04:47AM View BBCode
Originally posted by Fulla
3rd down conv % is slightly higher than the NFL. I find it difficult to stop a team on 3rd & any distance. Both Beta & RMFL increased their 3rd down conv % in the second season.
4th down conv % is slightly lower than the NFL. My personal settings are more aggressive on 4th down than most NFL coaches.
[Edited on 7-6-2011 by Fulla]
You say your 4th down settings are more aggressive than NFL coaches, and I agree with that for "going for it" on your opponent's half, but how often do you go for it on 4th & inches/short on your own half (compared with NFL coaches)?
I can answer it for you: you can't go for it. That's a major flaw in my eyes.
Fulla
Going for it on 4th & inches in own half could be used as a tanking tool. Sim owners IMO, will always be more aggressive than NFL coaches (no media scrutiny in SD).
Thanks KLKRTR for all the effort you put into this thread. Thanks Chris for all you put into the game.
July 06, 2011 at 11:50AM View BBCode
You can go for it late in each half anywhere on the field by using advanced prefs kicking/4th down situations.Going for it on 4th & inches in own half could be used as a tanking tool. Sim owners IMO, will always be more aggressive than NFL coaches (no media scrutiny in SD).
Thanks KLKRTR for all the effort you put into this thread. Thanks Chris for all you put into the game.
KLKRTR
Though, I'm pretty sure I could tank without having that option...
July 06, 2011 at 09:37PM View BBCode
My pleasure!Though, I'm pretty sure I could tank without having that option...
Pages: 1