2013 Sim results

Goodell
Posts: 3780
Joined: Mon Sep 15, 2008 3:44 am

Re: 2013 Sim results

Post by Goodell »

RebelFan wrote:One interesting thing to look at on that 49 yard play by the RB, is whether he should actually get it in the sim with such a dramatic disadvantage in the trenches.
Another thing I think we should take a look at is removing the .15 cap on grade disadvantages so that the full difference in opposing units can be realized in the sim.
Ie - 7.8 F7 vs. 5.6 oline would get .22 advantage rather than .15 max.
Not sure if that has ever been discussed before but I'm for it.
They are things to think about for potential future rule changes, but why they are the way they are now...

- We didn't have big plays originally. That data wasn't part of the update. It was just pure averages for a couple seasons. Then we added the big play data. We didn't adjust the big play figure for grades because it was felt it reflected bringing in the individual talent to some degree (which was why we added it). Maybe Adrian Peterson is on a team with mediocre talent around him (or some youngster with game breaking speed), but if they make a big play in reality it's likely due to their INDIVIDUAL game breaking ability. We don't want to be grades are the only thing that matters. Individual talent matters too. If you have an elite game breaker, you don't lose his speed just because he's on a bad team. Adrian Peterson's on a bad team. If he breaks a TD run for 80 yards in reality, we don't say our sim team can't have that if they have bad grades. His team has bad grades in reality too. He has the individual big play talent to bust loose so we respect that individual talent. Grades matter a lot in this game, but we give individuals their play making ability also, for at least that one play a game we have data for longest play.

- We cap the maximum advantages for both rushing and passing because of the extreme nature of our teams and the aim to have realistic outcomes. We tend to have teams loaded up to win now with huge grades and the other extreme of teams with horrible grades trying to tank it. We don't want unrealistic 120-0 scores, which could happen if we don't cap things. The idea is it's capped high enough so that grades really matter to the outcome (as I think they do), but also don't push it too far to super unrealistic proportions of scores and individual stats to where it's silly stats super inflated by grade advantages pushing individually far far far beyond their real stats. We want to keep them somewhat AROUND their real stats within some bump up or down based upon different sim supporting casts, but not so much that the grades matter more than real results. We want a mix of both mattering. If it wasn't capped then any team with great grades could NEVER have any bad games ever. That's unrealistic also, even if desired by teams with good grades. Individual performances from week to week matter also by design. Poor play can't be totally overcome by just buying great grades. They can help or hurt some, but real performance is key. That's why it's part simulation and part fantasy (fanulation). It's not ignore real performance and we'll simulate a completely different outcome totally independently. We want reality. What we see in the NFL we want to see here. If a player struggles badly in NFL, we want them to struggle here to some degree (plus or minus some range based upon supporting staff difference). We want that unpredictability of real life performance mattering from week to week, not that the season is predetermined before the NFL games start based upon who has the best grades. I hate that about the baseball sim leagues I'm in with their pre-season projections everything's based on ahead of time, and set out to have reality matter in the simulation here as a season develops with it's real happenings mattering. An unexpected player emerges and it matters. A former stud stinks it up in reality and it matters. Reality matters and keeps everyone in the game even if they didn't think they'd have the best team to start the year because things happen. Even rebuilding teams can sometimes jump into the playoff picture unexpectedly if a rookie emerges as an elite talent. Andrew Luck takes a 2-win team to the playoff the next year in reality too. Real individual talent emerging needs to matter here too by design, not just the grades before the season starts.

Overall, those things are there for balance. To have some alternative environment (different roster grades) and some real individual result both mattering. To have some checks to maintain that desired realism. There aren't 95-0 NFL games. I don't think we want a handful of QBs throwing 100% of their passes each week because we don't cap how much we can increase their percentage. I don't think we want Adrian Peterson rushing for 2 or 3 times his real numbers going for an insane 4,000 yards and 10+ yards every carry for an entire year if his sim team happens to built the best sim line ever. We don't want silly space outcomes. We don't want to be an arcade game. We want to be fairly real, as if we were in NFL front offices for real with realistic outcomes. We want things that look like the NFL we see each Sunday. Our teams are more extreme between the good and the bad, but there are the haves/havenots in NFL also. We'll be a little more extreme, but don't want to be a lot. We want to be realistic more than anything.
Official Statement from the Commissioner's Office
RebelFan
Posts: 469
Joined: Fri Apr 10, 2009 5:01 pm
Location: MS

Re: 2013 Sim results

Post by RebelFan »

Thanks commish for replying and taking the time.
I understand the desire for parity and that it's by design that the sim output not be all about the grades.
That being said, (and this is the part where I whine a little) I just don't think a team with no olinemen rated over 77 should be able to function to the tune of 4 scores against a defensive front where the lowest starter is an 82. Ever, really.
I think this SF defense is the best that has ever been assembled since I've been in fangm. It almost feels like being punished for being for having high ratings.
Otherwise this game is a lot more fantasy-like than I thought and the reasoning & idea behind a dominant defense / front 7 is greatly diminished.
And to reduce that defensive advantage & cap it off to make the score look a certain way is a bit disheartening to a GM doing all he can to win games.
Honestly, when I looked this matchup over beforehand, I was picturing something like that Monday night game a couple years ago when the Packers sacked Cutler about 12 times, because that's the kind of oline we're talking about here. Levi Brown & Parnell the starting tackles for goodness sake.
I feel a lot better after getting that off my chest. And I hope commish doesn't think I'm being too abrasive or that this is just sour grapes.
As a GM, the more I know about how the sim works the better informed I can be to build a successful team.

One thing I would add to the Peterson argument of bad team/great player. He has always had pretty good oline play, no matter how crummy the rest of his team may have been.
GM - Chicago Bears - AFFL
GM - San Francisco 49ers - DFFL

"Talent Hoarder"
Goodell
Posts: 3780
Joined: Mon Sep 15, 2008 3:44 am

Re: 2013 Sim results

Post by Goodell »

No problem, discussions like this are good overall.

I don't play EA Sports arcade games, but my understanding of it is that grades matter most. They don't pull in real performance stats I don't think from week to week. It's all about a player's grade. It's all 100% simulated and based only upon players grades given to them ahead of time (and the video game player's skill moving those around, but if video players equal then grades matter most).

In fantasy grades matter nothing. Only thing that matters is play on that given day. Could have a Hall of Fame QB but if he throws 3 INTs and 0 TDs his talent/grade doesn't matter. Only the stats on that day.

So those are two different extremes. Fantasy and Simulation. One grades are the only thing that matters and the other performance on a week-to-week basis is the only thing that matters.

We set out to find a middle ground there because in my opinion BOTH matter. A player's talent (grade) and his teammates grades should matter. How a player does on a given Sunday should also matter. We're not fantasy. We're not simulation. We're "fanulation" as I called it when started this idea. Entirely based upon a MIXTURE of both, not that grades were the only thing that mattered or that fantasy stats were the only thing that mattered -- but both mattered in balance.

Small samples are dangerous. As I said, people want upsets so you're going to get them. Had it been my way and we ran mini-series before the generator passed out results, maybe your record (and others) is different. But people want upsets so they are going to happen. That makes over-reacting to small samples difficult. We'll see in the big picture much clearer.

Over time in those big pictures, high grades have mattered tremendously to league champions over the course of 17 games and through the playoffs against stacked teams. That said, upsets can still happen and the majority of our members want the possibility of upsets. Having high grades isn't a disadvantage if looking historically or by game logic. One disadvantage can be turnovers and field position that is huge in play-by-play movement but often ignored on stats analysis of outcomes. And the biggest disadvantage for favored teams is that upsets can and will happen, and most of the league voted to not reduce the chances of upsets.

We can have legitimate discussions in off-season rule changes about moving the maximum caps up or down. If you have no caps, you get stupid results in our extreme team environment. I've ran those numbers and they go against my aim for realism when there are several 95-0 games and 900 yard passing nights with 10 TDs with 5 QBs throwing 100% completions and 5 RBs running for 400 yards a game. Put AP on a high grade team against a bad team and his 20 runs for 100 yard game turns into 200 yard game and 95-0 score with him picking up first downs on every single play all game long. Put same AP on low grade team against a top team and his 100 yard game becomes a horrible 20 rushes for 10 yard game despite him playing well in reality. It gets far from realism fast when you totally explode the averages so drastically when uncapped. That's making grades count way too much in my opinion and completely ignoring player talent and value to their individual skills.

On every team (even the really bad teams in NFL) there's probably at least one playmaker still who can bust out an athletic long play from time to time. No matter their teammate grades. Maybe they're a high pick athlete with blazing speed. If they manage to bust out a long play in reality (even from a bad team with bad supporting grades) we recognize that ability to have made that play with that individual talent and voted to apply those individual talents here also along with grades mattering a lot otherwise.

The aim of the game as setup wasn't to have players DRASTICALLY different from themselves in reality on a given Sunday. We want them very similar to how they play on Sunday, but just bumped up or down a bit depending upon other factors. We don't want completely different players than we're seeing in reality. We want to be real. Just a slightly modified reality because we have different rosters to where we can help or hurt these key players based upon our other moves. We can't make them entirely different people, but we can help/hurt them a bit within a reasonable range of realistic variation that still resembles what we see before us each week.

Some say grades should matter most. Some say performance should matter most. Here, though, we say there is a balance to be found between the two where both matter.
Official Statement from the Commissioner's Office
JonC
Posts: 341
Joined: Fri Jun 22, 2012 9:10 pm

Re: 2013 Sim results

Post by JonC »

This is a great read as a new GM here.

My personal opinion as a newbie...both grades and stats should matter, but stats seem to be the small sample while grades are the overall truth. I would definitely welcome discussing this at length in the off-season and have several ideas to throw into the fray.

Again, great read to start building a knowledge base.
DFFL Steelers GM: '13-'22
Regular Season Record: 77-85 (.475)
Division Championships: ’13, ’14, ’19
AFC WC Team: ’20

AFFL Bills GM: '20-?
Regular Season Record: 20-30 (.400)

BRFL Chargers GM: '21-?
Regular Season Record: 17-17 (.500)
AFC WC Team: '22
Strategist
Posts: 433
Joined: Tue Sep 01, 2009 11:06 pm

Re: 2013 Sim results

Post by Strategist »

Look I like the fact that upsets happen in our sim. That is REAL LIFE. Look at CLE beating CIN this week.
DFFL - DAL 09-20: 113-63 .642 (6-5) 3X DIV Champs. 6 Playoff apps. DFFL Bowl I Champs
CFFL - NYG 10-12: 34-13-1 .708
AFFL - WAS 13-19: 53-59 .473 (5-3) '14, '15, & '17 Div, '17 AFC Champs
FFFL - PIT 16-17: 45-19 .703 (3-3) '16-18 Div, 16' AFC Champs
jmdaz44
Posts: 65
Joined: Mon Apr 01, 2013 10:26 pm

Re: 2013 Sim results

Post by jmdaz44 »

I was looking through my team's results (CFFL HOU) this week, and late in the 4th quarter with my team up by 10, all my team was doing was throwing the ball. Shouldn't they have been mostly running the ball to kill the clock? Just curious if that's something we can influence or not.

Thanks!
CFFL Houston:
2013: 7-9
2014: 9-7
2015: 10-6 - Wildcard (0-1)
2016: 12-4 - AFC South Champ (0-1)
2017: 11-5 - AFC South Champ (1-1)
2018: 12-4 - AFC South Champ (1-1)
2019: 13-3 - Wildcard (4-0 - CFFL Champions)
Goodell
Posts: 3780
Joined: Mon Sep 15, 2008 3:44 am

Re: 2013 Sim results

Post by Goodell »

jmdaz44 wrote:I was looking through my team's results (CFFL HOU) this week, and late in the 4th quarter with my team up by 10, all my team was doing was throwing the ball. Shouldn't they have been mostly running the ball to kill the clock? Just curious if that's something we can influence or not.

Thanks!

If I'm looking at the right thing, for most of the 4th quarter up until the 2 minute warning it was just a one score game. It wouldn't go into hyper conservative run the ball on nearly every play mode early in the 4th with a one score game but would keep to the normal factors deciding whether plays run/pass per the game update inputs. Looks like all the RBs on the roster had as many carries or more than their game updates, so that would be part of the reason neutral calls in fairly close games wouldn't tilt more running if all the RBs already were fulfilling their updates. Had their been a huge RB update that it still hadn't fulfilled while a QBs update was, and the game situations weren't extreme where it was doing normal play calling mode then it would have had the game updates tilt more running to try to meet those stats.

Once the team got a FG later in the 4th to put a little more distance between it and it's no longer a one score game and it's late in the 4th, I believe it called only run plays after that to finish up the game.
Official Statement from the Commissioner's Office
jmdaz44
Posts: 65
Joined: Mon Apr 01, 2013 10:26 pm

Re: 2013 Sim results

Post by jmdaz44 »

Gotcha. I must have misread it the first time through because I thought I was up by two scores the entire 4th quarter for some reason. After re-reading it knowing I was only up by one score, it made a lot more sense. My bad. And you're right, once I did finally go up by two scores it just ran the ball the rest of the way. Thanks for responding so quickly. I wasn't overly concerned about the actual sim results, but was more interested in knowing if the sim takes situations like that into account.
CFFL Houston:
2013: 7-9
2014: 9-7
2015: 10-6 - Wildcard (0-1)
2016: 12-4 - AFC South Champ (0-1)
2017: 11-5 - AFC South Champ (1-1)
2018: 12-4 - AFC South Champ (1-1)
2019: 13-3 - Wildcard (4-0 - CFFL Champions)
TylerW
Posts: 283
Joined: Wed Jul 30, 2008 3:41 pm
Contact:

Re: 2013 Sim results

Post by TylerW »

1:44 | 2 and 10 @ 26 (74)... Peyton Manning (29-for-49 295 yds | 59%:59) passes complete to Darrel Young for 15 yards. TIMEOUT San Diego (1 left)
1:39 | 1 and 10 @ 11 (89)... Incomplete pass. Incomplete pass..
1:32 | 2 and 10 @ 11 (89)... Incomplete pass..
1:25 | 3 and 10 @ 11 (89)... Incomplete pass.
1:15 | 4 and 10 @ 11 (89)... Going for it on 4th down... incomplete pass.
Ball turned over on downs.
1:15 | 1 and 10 @ 11 (89)... LeVeon Bell runs for -1 yards (8-14yds | 1.8 ave:2.1). TIMEOUT San Diego (0 left)
1:08 | 2 and 11 @ 10 (90)... Robert Griffin III runs for 4 yards (9-66yds | 7.3 ave:8.6).
0:19 | 3 and 7 @ 14 (86)... Incomplete pass.
0:10 | 4 and 7 @ 14 (86)... Pat McAfee punts the ball 39 yards
Dexter McCluster returns the ball 8 yards.
0:00 | 1 and 10 @ 45 (55)...

End of Quarter 4 (Indianapolis 23 to San Diego 17)
This happened in the CFFL game.

First off, the fact that the Legend Peyton Manning throws 4 straight incomplete passes from the 11 in a row is completely unrealistic!!! ;) (Kidding, Troy - I know he already got his 2 TDs for the week and Indy had the advantage too)

But the real point of this post..I was out of TO's at 1:08 mark so how come RG3 was throwing the ball on 3rd and 17 with 19 seconds left when they could have just kneeled it? Didn't affect the outcome but I'm sure Zapotek would have been pissed (I would be too) if something would have glitched and I scored as a result of the punt.
GM SD Chargers CFFL
Franchise Record: 72-23-1 (Playoffs: 4-5)
-2008,2009,2010,2011,2013 AFC West Division Champions
-2008 AFC Champions
GM NY Giants AFFL
Franchise Record: 4-44
Post Reply