Page 1 of 4

Pre Off-Season Discussion Topics

Posted: Wed Jan 09, 2013 10:30 am
by tino38
1st topic:
Curious to see everyone else's reaction to the heavy passing league that we have turned into. In the AFFL there was 1 guy who threw for 5,000+ yards. 2 others were close to 5,000 but nothing out of the normal. In the CFFL 3 guys threw for 5,000+ yards with Tom Brady 8 yards shy of that. Joe Flacco has not thrown over 3,800 yards in his career until this season, yet this year he throws for 5,048. In the DFFL there were 3 guys that threw for 5,000+ yards. Tony Romo ended up passing the NFL passing record (5,476 set last year-Brees) by throwing for 5,697. The crazy thing here though is that not only did he pass the record, but Matt Ryan and Drew Brees also got very close to passing the record this season in the DFFL. Feels like we are going to have a 6,000 yard passer soon. Just kind of wondering if the passing should be turned down a bit.

2nd topic:
Last offseason we talked about making a move from Madden to ProFootballFocus. I am not opposed however I would still like to use Madden rating, as well as PFF. For instance, if we had the ability to decide we want to play like the 49ers and have a serious rushing attack with one if the best offensive lines, we could implement PFF stats to help build a solid rushing o line versus guys who are much better at pass blocking. And vice versa, take a team like the Packers where they have a very little rushing attack but one of the best passing games in the league, you then could find offensive lineman to fit that mold. Just my thoughts, look forward to see your guys thoughts as well.

Re: Pre Off-Season Discussion Topics

Posted: Wed Jan 09, 2013 11:46 am
by Goodell
Definitely open to ideas for the off-season discussions. We'll dive deeper into all that after the playoffs, but thanks for getting some ideas out there to start thinking about.

Some interesting ones there. We did make some changes because it seemed like we were way behind the passing curve in previous years. When first programmed 5-6 years ago or whatever, QBs weren't putting up the huge numbers they are in the pass happy NFL now, and felt like when crunching the numbers that we had to change with the times to keep up.

Might have been too much, though. I know it was too much for week 1, and I had to put in some reactionary tone-downs to avoid some of the higher extremes generated. Doing too much to try to hold stats down does bad things some times too, like sometimes having running plays late in the game when a team might otherwise be passing because the sim doesn't want to make the QB numbers too crazy beyond real attempts. So there's a tricky balance to find between being honest to the game situations and what a team should do versus trying to ride close to the game updates.

Some people complain about grades not mattering much anymore, but when QBs on elite grade teams get those performance boosts beyond their actual numbers because of great supporting casts and have much larger numbers than they do in reality as you point out, they matter. I don't think a QB thowing for more yards in the sim versus reality is automatically a bad thing (even many more yards) without looking at the context of are those QBs on elite grade teams where they received many QB performance boosts. If so, that improvement over real numbers is intended. We don't want everyone to be way over inflated numbers for all, though. We'd have to see the other side of things... are there QBs on poor supporting cast teams who had lower numbers than reality as intended? Are the QBs who were the most over on teams with a good running game also, or were they sim teams with a good QB and horrible RBs where the sim obviously is going to lean on the pass more than perhaps that QB's real more balanced reality. So some of the extreme situations might be perfectly understandable in context, but it's something to go deeper into and see if improvements can be made (probably, it's always a work in progress...)

I haven't had a chance to crunch all the numbers yet this year (and welcome anyone to assist there eventually over the long summer), but my guess is we probably are a bit too pass-oriented now.


Grades sources have come up before. I thought at one point one of the sources much discussed wasn't a free service and probably wouldn't like us publishing their paid service grades on public free site and causing trouble. Definitely open to suggestions there. I'm going to pursuit more business relationships ahead to make things more efficient (such as stats feeds instead of me collecting all the box scores manually) and it's possible some relationships with grading sources. Definitely open to thoughts there, but some of that might end up being dependent upon relationships built (if any) and services available at what costs or effort.

Re: Pre Off-Season Discussion Topics

Posted: Wed Jan 09, 2013 12:22 pm
by Troy S
My AFFL team might be a good case study for the run/pass balance with the two leading NFL MVP candidates Peyton Manning and Adrian Peterson on the same squad.

NFL Peyton Manning: 400-for-583 (68.6%) 4659 yards, 37 TD, 11 INT
SIM Peyton Manning: 400-for-550 (72.7%) 4831 yards, 44 TD, 10 INT

NFL Adrian Peterson: 348 rushes, 2097 yards, 12 TDs (40 receptions, 217 yards)
SIM Adrian Peterson: 270 rushes, 1878 yards, 19 TDs (31 receptions, 367 yards)

Some of this is pretty logically explained within the game intentions. I'd expect Manning to throw less on my sim team than he does in reality because he has Adrian Peterson (as well as his own real RBs Moreno/McGahee). I had a fairly week schedule and often had 10% to max 15% grade advantages many games which would account for increase in pass completion % (just over 4% more so not a ton more as it's already high in real numbers). Max performance boosts can reduce INT in update and yards/completion can be boosted throughout the game if TDs not met yet, etc. so less INT and more TD also makes sense but not too far off from reality. He got more yards in the sim with less attempts. My team doesn't have an elite WR really beyond what he has in reality, so that's something that should probably be checked into a bit -- put possibly explained by having a weak schedule and performance boosts often that boost those numbers throughout a game if hasn't reached game update milestones yet.

AP's fewer rush attempts also makes sense having Peyton Manning versus his real team situation that leaned on him so much in reality with young less effective QB. I believe I also had AP as RB2 often to open up the season with him coming off injury last year. He had more TDS in the sim, but I think that's logically explained also by having Peyton Manning and a weak schedule being in the red zone so often. He wouldn't get automatic TDS beyond his real TDS, but with a very high average per rush once Peyton gets the ball down to the goal line and doesn't have any more pass TDS in his update, the sim team often would understandably hand off instead of racking up more Peyton TDs and have AP's high average carry him over the goal. He had less receptions for more yards, but also I think understandably with Peyton's high yardage being the deciding factor mostly on how far Peyton's receptions go in yardage so Peterson benefiting there from better QB.

So some intended results there within designed reason I think for the most part overall, but at least in this case probably another case for needing to dial down yards on passes a little.

The closest AP came to real rush yardage was in CFFL (327 rushes for 1,994 yds) where his QB was Ponder also and was on a more similar team to his real circumstances perhaps. That makes sense. With a pretty good OL on that CFFL team, though, without checking into how many games AP had performance boosts, you'd think while close to reality might be a case where things leaned a little more passing and held him down a little where he had about 20 less rushes (not sure about how game situations may have lead to more pass calls if behind more than ahead, etc.).

At least in this case study, we're not way far off and where we differ in many cases seems logically by design with real vs. sim differences, but also suggests that basic feeling expressed of perhaps a little too pass friendly and some tweaks there to perfect that balance still needing to be made ahead for next season.

Re: Pre Off-Season Discussion Topics

Posted: Wed Jan 09, 2013 12:37 pm
by Troy S
And while yards per completion seem high overall, odd that Calvin Johnson set records in reality with 1,964 yards receiving but he didn't crack that total in any of the three sim leagues despite strong passing numbers overall. Some of that I think because he had different SIM QBs in some cases and others with sim receivers with better stats than real teammates sharing the receptions more. In a case where he had same SIM/NFL QB, despite our sim league being pass happier his sim QB version had fewer attempts also than reality where that QB set an NFL record for attempts this year.

Our sim had different guys at times more moving toward those record setting attempts and in more frequency, but perhaps also on teams/situations that would make passing more reasonable in that simulated environment and somewhat within reason if NFL players attempting in similar amounts in similar situations.

Re: Pre Off-Season Discussion Topics

Posted: Wed Jan 09, 2013 2:40 pm
by tino38
Yeah you may be entirely right and due to some teams being so heavily stacked with offensive weapons there may really not be anything we can do because in all actuality if some of these teams existed in real life NFL it's very possible these guys could put up numbers similar. I just thought it was a bit odd that some of the passing was increased by over 500 yards in some cases. I also do agree, I find it weird that not in a single league did Calvin Johnson break surpass or come really close to the WR yardage record in our sim. However in the DFFL it should be noted that NO plate without an actual starting QB and that will have played a major role in that towards his stats.

Your stats are very close to reasonable IMO and maybe it can be done to look deeper into the stats and ratings to see if it can be equaled on other teams as well. The AFFL NFC seems to be pretty loaded with good teams to and that could have had an impact towards keeping a guys stats closer to reality.

Now that you mention it above I do recall having that discussion about not being able to use their statistical findings because of they out their data out their for money rather than like Madden does for weekly Madden updates. There will probably not be a whole lot that can be done with that, then but glad I asked about it because I did not think about that.

Like I said, just wanted to open up some discussion, and I am sure as playoffs come to an end their will be quite a lot more things from everyone else. Thanks for the quick reply!

Re: Pre Off-Season Discussion Topics

Posted: Wed Jan 09, 2013 3:30 pm
by Ulrich82
I don't think the 3 5,000 yard passers in the CFFL is that big of an issue. There was only 1 in the NFL this season (with Stafford falling about 50 yards short), but the NFL had 3 last season with Eli coming within 100 yards of it. Clearly, there is precedent for such a number in the modern NFL offense. Some statistical fluctuations are to be expected.

In defense of Flacco, I added Andre Johnson this year which along with Dex Bryant's improvement gave me the top WR corps in the CFFL this season. I only have an average offensive line, but I think the stats generally make sense. In the NFL this season, Flacco completed 59.7% of his throws with a yards per attempt of 7.18. In CFFL, he completed 64.4% of his passes with a yards per attempt of 7.71 (Interestingly enough, in both NFL and CFFL, he had almost exactly 12 yards per completion, but perhaps that is just coincidence).

I think the better argument might be that we have a clock problem IF we have too many plays per game. In the CFFL, I count 9 teams with a QB who attempted more than 700 passes. In the past 3 seasons, only 2012 Mathew Stafford had over 700 attempts in the NFL (727), and the cap usually appears to be ~650-670. However, there does seem to be an upward trend over the past few years. If Flacco didn't have 655 attempts in CFFL (he had 531 in the NFL), he never would have sniffed 5000 yards. Of course, the sim needs to find a way to run plays regardless of the number of attempts the NFL players in bring in. Looking at my team in particular, I ran 412 running plays and 655 passsing plays (not counting QB runs and sacks I believe). This is equivalent to about 66.7 offensive plays per game. Looking at the 2012 numbers for teams offensive plays per game, this would be exactly equivalent to 7th most in the 2012 NFL (with the top being NE with 74.4 plays per game). Without looking at the rest of the CFFL though, it is hard to say if we really average significantly higher plays per game than the NFL.

Re: Pre Off-Season Discussion Topics

Posted: Wed Jan 09, 2013 3:48 pm
by tino38
You may be onto something there Ulrich. The pass attempts may be more of the issue than the actual yardage. That is a crazy high amount of pass attempts so that may be where the issue lies. However, I don't think that just because of Dez and Andre that justifies the 5,000+ yards passing. Aaron Rodgers in real life has arguable the best WR core top to bottom with very good WR's on the field at all times, and on a team that isn't very good at running the ball. He has yet to even break 5,000 in his career. Flacco threw for 1,231 yards more than he did in real life. That's the widest margin I have seen between FANGM passing stats and real life stats. That's what is odd to me.

Re: Pre Off-Season Discussion Topics

Posted: Wed Jan 09, 2013 4:02 pm
by Goodell
I do think some tweaks needed, as we always will tweak a little.

But one thing to keep in mind with analysis of how much to tweak... we don't strive for equal stats.

We want to have a game where you have Player X at QB and get to put together a team building GM game plan for how to improve with Player X as your QB by getting him more personnel help.

Now we have limits on how much a vastly improved or vastly inferior supporting cast will allow Player X's stats to shift, but the aim would be having avenues for our FanGMs to have a QB with better sim stats than real stats IF they've done a great job of building up a strong offensive line for him and strong receivers to where that QB has a dominant sim offense going up against weaker sim defenses and logically giving him a boost under those advantageous situations.

Fantasy sports is what actually happens and points for actual events. Here we don't want actual, but what COULD or SHOULD happen under different circumstances. Imagine if this QB had 2 more all-pro offensive lineman and 2 more elite receivers... that's what we want to try to simulate in a realistic manner giving our sim GMs a chance to alter reality to their advantage.

Overall, not all teams in the league should be able to boost up their QB/RB stats. The really superior offensive lines and supporting casts should (as that's the incentive we want encouraging people to build those supporting casts as a very important aspect of winning), but there should be teams on the other side of the stick who go up against those superior teams and should have negative impacts on their offensive player stats.

We want to ensure we don't get unrealistically crazy in how much we have grades boost up stats, but we do want to make sure that there is a way to help Player X do better in the sim than reality.

Too many plays was mentioned in one of the posts. I came to the conclusion last year that we had too few plays per game compared to the high powered offenses in reality and shifted some of that. It seemed our opportunities were too few for many elite players to achieve in the sim what they were in reality, so tried to add a little more plays. Maybe too much and might need slight adjustment to less plays, as well as looking again at play call mix. The real challenge in that, though, will always be some players on real teams will put up HUGE amounts of plays with hurry up offenses that stay on the field while other teams will be more traditional. Right now the time per play is fairly universal for all, and just speeds up in cases where a team might be doing hurry up based upon game situations. May look at a more dynamic clock that also factors in how players are doing against their averages to where if they are way below their attempts that their offense hurries up to allow more attempts (if they are successful at holding onto the ball), but there are other factors to that also which could impact game play such as if a team was ahead and wanting to slow the clock down wouldn't want the sim speeding it up on them just to gather more stats where winning the game should be more important than trying to get more stats. But a couple things to look at there. Probably the first thing I'll do is keep it simple and just add a little bit more time per play on the whole to where it's still faster than previous years but not quite as fast as this year. It's just a couple of seconds difference per play within a random range on each play, but over a whole season does add up. So we monitor and keep re-adjusting.

Re: Pre Off-Season Discussion Topics

Posted: Wed Jan 09, 2013 4:35 pm
by tino38
I do agree that a guy should be able to go above and beyond regular stats when having elite talent around him as weapons and offensive line, just shocked to see the amount that Flacco increased by. Might just be a non issue.

Re: Pre Off-Season Discussion Topics

Posted: Thu Jan 10, 2013 12:37 am
by Ulrich82
But you can't cap Flacco's yardage. Based on the info in the game recaps, I believe the sim largely relies on completion percentage and yards per attempt. In both those categories, CFFL Flacco had moderate increases over NFL Flacco over the whole season. Is was the total plays that led to such high total yardage.

If the sim situation calls for a pass, it isn't fair to make my team run just because Flacco only attempted 20 NFL passes that week. Similarly, his yards per attempt shouldn't be negatively impacted to fit some total yardage cap.