 |
View Old Forum Thread
Old Forum Index » Football Beta Testing » Football Beta Test Discussion » Two-point conversions
Admin
--Chris
Two-point conversions
March 21, 2011 at 10:06PM View BBCode
I found some AWESOME data on two point conversions that makes the strategy setting for it as simple as putting in a slider for "conservative" or "aggressive". I'm trying to adapt it to the sim now; we might have them in the game later tonight.--Chris
redcped
March 21, 2011 at 11:57PM View BBCode
Excellent. This will also affect some other strategies, like timeouts.
Admin
Right now, it goes all the way from "Never go for 2" to "Always go for 2" but that makes the only really useful settings clustered around the middle of the slider. I'm thinking that maybe increasing or decreasing the chance of a 2-pointer by 50% maximum might be more useful.
Essentially, the data I have basically says for each possible time and score difference combo, there's a certain percentage at which there is a break-even difference in win probability, so with the average NFL team converting 40% of the time, a value of 50 (the middle) on the slider would correspond to going for 2 whenever the chart had a value of 40% or less. I'm thinking that rather than put the top and bottom of the slider at 100% and 0%, I'll put them at 60% and 20% instead.
For example, let's look at the situation where there is less than 90 seconds left on the clock and after a touchdown you are down by 1. A kick nearly assures you of the tie, the 2 point will give you a win or a loss. The chart says to go for a 2-point conversion if you are at least 60% sure of making it; if I have the sliders set to 60% and 20%, you'd have to be maxed out on "Aggressive" to go for it in that situation, whereas the slider would only be up a third of the way from center to do that if I use 0% and 100% So the question is: Do you feel like going for 2 when down by 1 with 90 seconds on the clock is a "very aggressive" move or a "somewhat aggressive" move?
--Chris
[Edited on 3-22-2011 by Admin]
March 22, 2011 at 12:44AM View BBCode
One question I have is "How far do we let people adjust the settings?"Right now, it goes all the way from "Never go for 2" to "Always go for 2" but that makes the only really useful settings clustered around the middle of the slider. I'm thinking that maybe increasing or decreasing the chance of a 2-pointer by 50% maximum might be more useful.
Essentially, the data I have basically says for each possible time and score difference combo, there's a certain percentage at which there is a break-even difference in win probability, so with the average NFL team converting 40% of the time, a value of 50 (the middle) on the slider would correspond to going for 2 whenever the chart had a value of 40% or less. I'm thinking that rather than put the top and bottom of the slider at 100% and 0%, I'll put them at 60% and 20% instead.
For example, let's look at the situation where there is less than 90 seconds left on the clock and after a touchdown you are down by 1. A kick nearly assures you of the tie, the 2 point will give you a win or a loss. The chart says to go for a 2-point conversion if you are at least 60% sure of making it; if I have the sliders set to 60% and 20%, you'd have to be maxed out on "Aggressive" to go for it in that situation, whereas the slider would only be up a third of the way from center to do that if I use 0% and 100% So the question is: Do you feel like going for 2 when down by 1 with 90 seconds on the clock is a "very aggressive" move or a "somewhat aggressive" move?
--Chris
[Edited on 3-22-2011 by Admin]
Admin
(Note: I can set the top and bottom of the "aggressive/conservative" scale independently.)
--Chris
March 22, 2011 at 01:02AM View BBCode
An example from the other end of the spectrum: A touchdown with 9 minutes left in the 4th quarter leaves you up by 1. The win probability chart says you should go for 2 if you feel like you even have a 21% chance of converting. If I have the sliders set at 60% and 20%, having the slider down at minimum means you will take the sure thing rather than go for 2, which would mean that an opponent's field goal would only tie rather than win. Is not going for 2 in this case being extremely conservative, or only moderately conservative?(Note: I can set the top and bottom of the "aggressive/conservative" scale independently.)
--Chris
Hamilton2
March 22, 2011 at 01:04AM View BBCode
I think that the 60/20 numbers are good, and I really like the approach that you are taking with this.
Admin
Right now, I'm thinking of setting the default down around 32%, so the bottom half of the slider would go from 32% to 20% and the top half would go from 32% to about 65%, unless the answers to the questions I've posted are wildly different. The minimum and maximum settings should be, I think, on borderline "You're nuts".
--Chris
March 22, 2011 at 02:07AM View BBCode
40% would be the statistical average, but would result in probably double the number of 2-pt attempts compared to real life, just because NFL coaches are so conservative. The defaults should probably reflect real life, not the ideal. (Otherwise we'd go for it far more on 4th down and almost never punt.)Right now, I'm thinking of setting the default down around 32%, so the bottom half of the slider would go from 32% to 20% and the top half would go from 32% to about 65%, unless the answers to the questions I've posted are wildly different. The minimum and maximum settings should be, I think, on borderline "You're nuts".
--Chris
redcped
http://www.theredzone.org/Features/TwoPointConversionChart.aspx
This, however, is probably much more useful for coding:
http://www.footballcommentary.com/twoptchart.htm
Finally, this might be useful because the bottom chart breaks it down by expected remaining possessions in the game. Shouldn't be too hard to determine that average, factoring in the faster drives in the 2-minute game:
http://fengnet.com/book/Statistics.Hacks/I-0596101643-CHP-5-SECT-9.html
[Edited on 3-22-2011 by redcped]
[Edited on 3-22-2011 by redcped]
March 22, 2011 at 03:53AM View BBCode
Here's the classically accepted chart for what to do in what situation:http://www.theredzone.org/Features/TwoPointConversionChart.aspx
This, however, is probably much more useful for coding:
http://www.footballcommentary.com/twoptchart.htm
Finally, this might be useful because the bottom chart breaks it down by expected remaining possessions in the game. Shouldn't be too hard to determine that average, factoring in the faster drives in the 2-minute game:
http://fengnet.com/book/Statistics.Hacks/I-0596101643-CHP-5-SECT-9.html
[Edited on 3-22-2011 by redcped]
[Edited on 3-22-2011 by redcped]
Admin
--Chris
March 22, 2011 at 04:51AM View BBCode
The footballcommentary chart is the one I am using, although I've interpolated values down to 90-second increments and extended it all the way to the beginning of the game.--Chris
Admin
--Chris
March 22, 2011 at 07:08AM View BBCode
I've uploaded a sample to Beta... this is only a sample, it will not save your setting yet. The slider is at the bottom of the Kicking setiings, and it will show you some examples of what you would do in certain situations as you move the slider. Please let me know what you think.--Chris
Admin
(Note: The table seen on this page is a subset of the full data, which runs from -30 to +30 and from 90 seconds to 60 minutes in 90 second increments; it just shows a subset to give you a feel of what the setting represents.)
--Chris
[Edited on 3-22-2011 by Admin]
March 22, 2011 at 07:37AM View BBCode
I'm fairly pleased with how that came out. I'm kinda feeling like the setting currently marked as 90 should probably be the maximum... any thoughts?(Note: The table seen on this page is a subset of the full data, which runs from -30 to +30 and from 90 seconds to 60 minutes in 90 second increments; it just shows a subset to give you a feel of what the setting represents.)
--Chris
[Edited on 3-22-2011 by Admin]
Fulla
Does it use the goal line offense strategy?
[Edited on 3-22-2011 by Fulla]
March 22, 2011 at 08:15AM View BBCode
Wow that is really cool.Does it use the goal line offense strategy?
[Edited on 3-22-2011 by Fulla]
Admin
--Chris
March 22, 2011 at 08:18AM View BBCode
Yes. I'm not sure when you looked at it, but at some point I added the line "Both offense and defense will use their 4th and short Goal Line settings in a 2 point conversion attempt.". It makes the most sense, I think, rather than building a whole extra set of settings.--Chris
Fulla
March 22, 2011 at 08:43AM View formatted
You are viewing the raw post code; this allows you to copy a message with BBCode formatting intact.
I dont see how people could abuse this. At first, out of curiosity, they will test the limits. The success rate will eventually make them use caution.
Admin
It is worth noting that private leagues will have the option of turning off 2-point conversions entirely.
--Chris
March 22, 2011 at 08:57AM View BBCode
The worst case for abuse would be trying to tank, I suppose.It is worth noting that private leagues will have the option of turning off 2-point conversions entirely.
--Chris
Admin
--Chris
March 22, 2011 at 09:26AM View BBCode
OK, you should be able to save these settings now.--Chris
redcped
March 22, 2011 at 04:52PM View BBCode
So would it be useful to try maximum aggressiveness to see if the conversion rates are where we want them?
Admin
--Chris
March 22, 2011 at 05:02PM View BBCode
Well, i still have to build the actual part in the game... all I've finished is the interface and strategy loaders. :) Working on the gameside now.--Chris
redcped
OK, cool.
I think the interface works well. It's easy to understand and maneuver.
March 22, 2011 at 05:06PM View BBCode
Originally posted by Admin
Well, i still have to build the actual part in the game... all I've finished is the interface and strategy loaders. :) Working on the gameside now.
--Chris
OK, cool.
I think the interface works well. It's easy to understand and maneuver.
Admin
The scoring summary is broken, but otherwise the game appears to flow properly. Right now I'm just testing the mechanics, not the strategy, so I have all conversions as 2-point attempts right now. Got some work to do to fix the scoring summary and check for all the possible flaws, but I may be able to turn this loose on Beta tonight.
--Chris
March 22, 2011 at 08:58PM View BBCode
In my development league I just had the first successful 2 point conversion, and several failed ones. :)The scoring summary is broken, but otherwise the game appears to flow properly. Right now I'm just testing the mechanics, not the strategy, so I have all conversions as 2-point attempts right now. Got some work to do to fix the scoring summary and check for all the possible flaws, but I may be able to turn this loose on Beta tonight.
--Chris
Admin
--Chris
March 22, 2011 at 09:24PM View BBCode
In testing so far (small smaple size), 2 point conversions are successful 43% of the time.--Chris
Fulla
I personally plan to mix it up a bit
[Edited on 3-23-2011 by Fulla]
March 23, 2011 at 12:56AM View BBCode
Is the testing using the default 4th down & short for both offense and defense? Or is it testing various plays? I ask this because probably 99% of owners will be using default 4th & short.I personally plan to mix it up a bit
[Edited on 3-23-2011 by Fulla]
Admin
I'm also putting the defensive stats onto the stats pages, although a lot of the numbers are still wrong or missing, and testing the changes to the O-line. All this may get moved to Beta in the wee hours, just depends on how long I can stay awake and if I can find the bugs. :)
--Chris
[Edited on 3-23-2011 by Admin]
March 23, 2011 at 05:06AM View BBCode
Right now it's using the defaults. It's 99% ready to move to Beta; unfortunately I'm having a weird issue where if the touchdown is by the defense, the offense goes for the 2-point conversion. :oI'm also putting the defensive stats onto the stats pages, although a lot of the numbers are still wrong or missing, and testing the changes to the O-line. All this may get moved to Beta in the wee hours, just depends on how long I can stay awake and if I can find the bugs. :)
--Chris
[Edited on 3-23-2011 by Admin]
Admin
--Chris
March 23, 2011 at 07:52PM View BBCode
The code is up, and a quarter has been played with it. There were three touchdowns in this quarter; one in the Boston game (they kicked) and two by me (OAK), both of which were two point attempts, both of which failed.--Chris
Fulla
March 24, 2011 at 02:24AM View BBCode
The Newballers tried a 4th quarter two pt pass and it was incomplete. My 2 pt aggression is set at 72%.... Problem is, I can't score any touchdowns.Pages: 1 2