Stat Accuracy

hfield007 · 01-31-2014, 06:25 AM

I hope this will be handled better than FOF.

One problem I constantly ran into was the rushing record being broke a hundred times over and watching runningbacks consistently rack up 1800 yard seasons for ten years.

I am hoping in BTS that runningbacks will age differently from the other positions giving us realistic results like one year wonder RBs and the very rare ten year workhorse. I am also hoping that just like in OOTP that potentials can fluctuate year to year.

But mostly hoping that each position ages differently just like in RL.

Markus Heinsohn · 01-31-2014, 06:28 AM

Just like in OOTP, we'll definitely make sure stats are very accurate and player development behaves properly so that stats are still accurate 100 years down the road. This is one of the top priorities for the game

Kobeck · 01-31-2014, 06:36 AM

Qrashman · 01-31-2014, 08:03 AM

This company was built off ootp, a baseball game in a community where statistical accuracy means everything. I think they'll get this part right.

deanver · 01-31-2014, 08:15 AM

Agreed 100% I LOVE stats.

Francis Cole · 01-31-2014, 09:10 AM

I think there are two different things tied up in the question we have to look at:

1)Are Stats accurate
2)How is the player progression/regression.

Getting the stats accurate is the main focus of the 2D Engine, but it is also the toughest to do. The avg Yd per Rush last season was a tad over 4.0
Now if the user simply chooses to run all the time, they will find that their average will be far less than 4.0 (as defences know they will run all the time).
So we have to strike the balance between getting accurate stats, but at the same time making sure the AI reacts to the opposition correctly.

Secondly, player progression; This is something the health map was designed for! The fatigue on the knees of a a RB who gets 1800+ for 10 years will be extreme. I doubt he'd even get to 10 years with that kind of wear. Player injuries will play a crucial role.

bigtrouble77 · 01-31-2014, 09:32 AM

FOF7 was such a massive letdown. The news of BTS really makes that disappointment easier to take.

There are aspects of FOF that I absolutely love (stats being one of them), but the latest release really failed to capitalize on (fixing) all of the things that makes that game mediocre.

Stats are the last thing I'd criticize with that game. Jim seems to build that game for himself and only himself. It's very evident that he refuses to fix critical use ability issues simply because he doesn't care. He doesn't even have a support email published.

Anyway, looking forward to this BTS.

bigtrouble77 · 01-31-2014, 09:41 AM

Most sports have trends (usually based on rule changes), but I've never seen a sports sim support something like that. That would be a really interesting dynamic component to have... possibly have the ability to vote yearly on minor rule changes that could create interesting trends. Perhaps have the AI actively try to optimize for these types of changes.

I'd love to have a dynamic component like that rather than obsessing over whether the current stats perfectly reflect what the NFL is producing this year.

Quote:

Originally Posted by Francis Cole

I think there are two different things tied up in the question we have to look at:

1)Are Stats accurate
2)How is the player progression/regression.

Getting the stats accurate is the main focus of the 2D Engine, but it is also the toughest to do. The avg Yd per Rush last season was a tad over 4.0
Now if the user simply chooses to run all the time, they will find that their average will be far less than 4.0 (as defences know they will run all the time).
So we have to strike the balance between getting accurate stats, but at the same time making sure the AI reacts to the opposition correctly.

Secondly, player progression; This is something the health map was designed for! The fatigue on the knees of a a RB who gets 1800+ for 10 years will be extreme. I doubt he'd even get to 10 years with that kind of ware. Player injuries will play a crucial role.

Michael V. · 01-31-2014, 11:13 AM

Quote:

Originally Posted by Qrashman

This company was built off ootp, a baseball game in a community where statistical accuracy means everything. I think they'll get this part right.

Agreed.

Ben E Lou · 01-31-2014, 11:51 AM

I have major concerns about the ability to do a statistically accurate 2D sim engine for football. Typically such an endeavor creates far too many "max-length" plays, resulting in shorter-gain plays having to be toned down to make averages come out right. Be careful about that.

Markus Heinsohn · 01-31-2014, 12:00 PM

Quote:

Originally Posted by SkyDog

I have major concerns about the ability to do a statistically accurate 2D sim engine for football. Typically such an endeavor creates far too many "max-length" plays, resulting in shorter-gain plays having to be toned down to make averages come out right. Be careful about that.

Francis has plenty of time to fine-tune his engine. I'm sure the results will be very good.

Francis Cole · 01-31-2014, 12:02 PM

Quote:

Originally Posted by SkyDog

I have major concerns about the ability to do a statistically accurate 2D sim engine for football. Typically such an endeavor creates far too many "max-length" plays, resulting in shorter-gain plays having to be toned down to make averages come out right. Be careful about that.

The good news is that the engine is already year+ in the works.

I've actually found the problem to be the opposite!!

adaher · 01-31-2014, 12:06 PM

Football is HARD to do right and I don't think anyone has yet come close, at least compared to the accuracy of baseball sims.

Plus there's SO much variation in player performance depending on the performance of his teammates, unlike baseball where individual performance is fairly easy to figure out. Put Babe Ruth in his prime on the Chicago Cubs and he's still the best player in the game.

But take Mark Rypien away from the Washington Redskins and he's an average QB at best. Likewise, you've got those Denver running backs where for years anyone could rush for over 1000 yards. Would Terrell Davis do as well on the Miami Dolphins?

It's probably less of an issue if you're using fictional players, or keeping players on their historical teams, but doing a historical-league type thing where players can go anywhere is going to be devlishly hard to get right. I salute you guys for trying.

Pacoheadley · 01-31-2014, 02:06 PM

Quote:

Originally Posted by adaher

Football is HARD to do right and I don't think anyone has yet come close, at least compared to the accuracy of baseball sims.

Plus there's SO much variation in player performance depending on the performance of his teammates, unlike baseball where individual performance is fairly easy to figure out. Put Babe Ruth in his prime on the Chicago Cubs and he's still the best player in the game.

But take Mark Rypien away from the Washington Redskins and he's an average QB at best. Likewise, you've got those Denver running backs where for years anyone could rush for over 1000 yards. Would Terrell Davis do as well on the Miami Dolphins?

It's probably less of an issue if you're using fictional players, or keeping players on their historical teams, but doing a historical-league type thing where players can go anywhere is going to be devlishly hard to get right. I salute you guys for trying.

For this reason I hope they find some way to make coaches extremely important without just giving away all the coaches ratings very easily. Some sort of hybrid with ratings and schemes and other hidden ratings possibly? Coaching is such an important part of football compared to baseball, but I know that this is a challenge Francis is up for

Alaina · 01-31-2014, 05:33 PM

Quote:

Originally Posted by adaher

Football is HARD to do right and I don't think anyone has yet come close, at least compared to the accuracy of baseball sims.

Plus there's SO much variation in player performance depending on the performance of his teammates, unlike baseball where individual performance is fairly easy to figure out. Put Babe Ruth in his prime on the Chicago Cubs and he's still the best player in the game.

But take Mark Rypien away from the Washington Redskins and he's an average QB at best. Likewise, you've got those Denver running backs where for years anyone could rush for over 1000 yards. Would Terrell Davis do as well on the Miami Dolphins?

It's probably less of an issue if you're using fictional players, or keeping players on their historical teams, but doing a historical-league type thing where players can go anywhere is going to be devlishly hard to get right. I salute you guys for trying.

I think this would make it important to coach every game, and even hold practices and test out things, seeing what situations would go best with your roster.

Kobeck · 01-31-2014, 09:47 PM

football is multi variable for sure

teams try and acquire guys that fit their system, well the good ones do - lol

as a for instance I am a Titan fan and they drafted 5-10 185 pound 4.24 Chris Johnson who by anyones defination is a scatback. But the Titans have just plugged him in as starter in their traditional power run scheme.
Although there has been some success statwise the Ws have not followed.

The Titans should have altered their O to take true advantage of CJs unique talents or else drafted someone else with that pick IMO

not sure how the AI should handle that

BaseballMan · 01-31-2014, 11:24 PM

Plus how do you rate a offensive guard from 1935? The 72 dolphins line was great but they werent as good before they came together in miami.
I just hope we dont see a qb passing for 55 tds in the 1960s. Dont think the ghost of Vince Lombardi would like it. Dont worry about pleasing the community just dont tick off the ghost of Vince Lombardi.
Actually i wish Francis all the luck and i hope we as a community will remember to have a little patience.

Francis Cole · 02-01-2014, 06:14 AM

Thank you BaseballMan

packersfan · 02-02-2014, 01:12 PM

I have no doubt that this game will have great stat accuracy and the ability to fine-tune it to reflect game-style changes. One problem with Madden besides trying to play ACCURATE and CHALLENGING GM style is that the stats system can not be tweaked to reflect gameplay changes. For example, in Madden 2012 most QBs don't surpass 4000 yards passing...obviously this doesn't really match real-life stats a few years later, not to mention RBs rushing for TOO many yards. Looking forward to a full-football game (with the exception of seeing the guys hitting each other).

Jose Cardenal's Afro · 02-02-2014, 03:22 PM

Smashed in between the cure for cancer and world peace was my wish for this game......

01-31-2014, 06:25 AM	#1
hfield007 All Star Starter Join Date: Jul 2009 Posts: 1,712	Stat Accuracy I hope this will be handled better than FOF. One problem I constantly ran into was the rushing record being broke a hundred times over and watching runningbacks consistently rack up 1800 yard seasons for ten years. I am hoping in BTS that runningbacks will age differently from the other positions giving us realistic results like one year wonder RBs and the very rare ten year workhorse. I am also hoping that just like in OOTP that potentials can fluctuate year to year. But mostly hoping that each position ages differently just like in RL.

01-31-2014, 06:28 AM	#2
Markus Heinsohn Developer OOTP Join Date: Dec 2001 Location: Germany Posts: 24,806	Just like in OOTP, we'll definitely make sure stats are very accurate and player development behaves properly so that stats are still accurate 100 years down the road. This is one of the top priorities for the game __________________ Buy Out of the Park Baseball 25 now! Download OOTP Go 25 for iOS Download OOTP Go 25 for Android

01-31-2014, 08:03 AM	#4
Qrashman All Star Reserve Join Date: Dec 2001 Location: NY Posts: 651	This company was built off ootp, a baseball game in a community where statistical accuracy means everything. I think they'll get this part right. __________________ ------------------- Montreal Expos / Calgary Storm, HBL HBL Baseball - Founded March 2001. HBL Waiting List

01-31-2014, 09:10 AM	#6
Francis Cole Lead Developer for BTS Join Date: Nov 2013 Posts: 854	I think there are two different things tied up in the question we have to look at: 1)Are Stats accurate 2)How is the player progression/regression. Getting the stats accurate is the main focus of the 2D Engine, but it is also the toughest to do. The avg Yd per Rush last season was a tad over 4.0 Now if the user simply chooses to run all the time, they will find that their average will be far less than 4.0 (as defences know they will run all the time). So we have to strike the balance between getting accurate stats, but at the same time making sure the AI reacts to the opposition correctly. Secondly, player progression; This is something the health map was designed for! The fatigue on the knees of a a RB who gets 1800+ for 10 years will be extreme. I doubt he'd even get to 10 years with that kind of wear. Player injuries will play a crucial role. Last edited by Francis Cole; 01-31-2014 at 10:32 AM.

01-31-2014, 09:32 AM	#7
bigtrouble77 Minors (Double A) Join Date: Nov 2003 Posts: 125	FOF7 was such a massive letdown. The news of BTS really makes that disappointment easier to take. There are aspects of FOF that I absolutely love (stats being one of them), but the latest release really failed to capitalize on (fixing) all of the things that makes that game mediocre. Stats are the last thing I'd criticize with that game. Jim seems to build that game for himself and only himself. It's very evident that he refuses to fix critical use ability issues simply because he doesn't care. He doesn't even have a support email published. Anyway, looking forward to this BTS. Last edited by bigtrouble77; 01-31-2014 at 09:43 AM.

01-31-2014, 06:36 AM	#3
Kobeck Major Leagues Join Date: Nov 2004 Posts: 491

01-31-2014, 08:15 AM	#5
deanver Bat Boy Join Date: Apr 2012 Posts: 2	Agreed 100% I LOVE stats.

01-31-2014, 11:51 AM	#10
Ben E Lou Front Office Football Central Join Date: Feb 2002 Location: Souf Cackilacky Posts: 1,762	I have major concerns about the ability to do a statistically accurate 2D sim engine for football. Typically such an endeavor creates far too many "max-length" plays, resulting in shorter-gain plays having to be toned down to make averages come out right. Be careful about that.

01-31-2014, 12:06 PM	#13
adaher Minors (Rookie Ball) Join Date: Apr 2005 Posts: 22	Football is HARD to do right and I don't think anyone has yet come close, at least compared to the accuracy of baseball sims. Plus there's SO much variation in player performance depending on the performance of his teammates, unlike baseball where individual performance is fairly easy to figure out. Put Babe Ruth in his prime on the Chicago Cubs and he's still the best player in the game. But take Mark Rypien away from the Washington Redskins and he's an average QB at best. Likewise, you've got those Denver running backs where for years anyone could rush for over 1000 yards. Would Terrell Davis do as well on the Miami Dolphins? It's probably less of an issue if you're using fictional players, or keeping players on their historical teams, but doing a historical-league type thing where players can go anywhere is going to be devlishly hard to get right. I salute you guys for trying.

01-31-2014, 09:47 PM	#16
Kobeck Major Leagues Join Date: Nov 2004 Posts: 491	football is multi variable for sure teams try and acquire guys that fit their system, well the good ones do - lol as a for instance I am a Titan fan and they drafted 5-10 185 pound 4.24 Chris Johnson who by anyones defination is a scatback. But the Titans have just plugged him in as starter in their traditional power run scheme. Although there has been some success statwise the Ws have not followed. The Titans should have altered their O to take true advantage of CJs unique talents or else drafted someone else with that pick IMO not sure how the AI should handle that

01-31-2014, 11:24 PM	#17
BaseballMan Hall Of Famer Join Date: Jan 2002 Posts: 6,404	Plus how do you rate a offensive guard from 1935? The 72 dolphins line was great but they werent as good before they came together in miami. I just hope we dont see a qb passing for 55 tds in the 1960s. Dont think the ghost of Vince Lombardi would like it. Dont worry about pleasing the community just dont tick off the ghost of Vince Lombardi. Actually i wish Francis all the luck and i hope we as a community will remember to have a little patience.

02-01-2014, 06:14 AM	#18
Francis Cole Lead Developer for BTS Join Date: Nov 2013 Posts: 854	Thank you BaseballMan

02-02-2014, 01:12 PM	#19
packersfan Minors (Rookie Ball) Join Date: Jun 2004 Posts: 24	I have no doubt that this game will have great stat accuracy and the ability to fine-tune it to reflect game-style changes. One problem with Madden besides trying to play ACCURATE and CHALLENGING GM style is that the stats system can not be tweaked to reflect gameplay changes. For example, in Madden 2012 most QBs don't surpass 4000 yards passing...obviously this doesn't really match real-life stats a few years later, not to mention RBs rushing for TOO many yards. Looking forward to a full-football game (with the exception of seeing the guys hitting each other).

02-02-2014, 03:22 PM	#20
Jose Cardenal's Afro Major Leagues Join Date: Apr 2007 Location: Limbo Posts: 374	Smashed in between the cure for cancer and world peace was my wish for this game......