|
Post by ezlee2 on Sept 27, 2018 18:14:32 GMT -5
Man, I hope he's doing okay. What a great guy and the ultimate ambassador for FOF. He was a huge inspiration for me as to how I'd approach the RZB. Those days were really special. I'll never forget the first time we really started to examine QB height with the two stud QBs in the draft that year. I had 1.1 and Fire Fly had 1.2. There was the prototype 6'5" 245 lb guy and then the ultimate Flutieque player what was 5'8" 195 lb guy. Both were surefire studs and both wound up winning bowls (UKPatfan won back to back with my squad after I stepped down)
Anyone remember their names? I for the life of me cannot.
|
|
|
Post by WilleB on Sept 27, 2018 18:23:05 GMT -5
Interested in why timing gets very little attention here. I'm huge on timing and accuracy. This guy is pretty darn good. His picks have decreases as our WR corps have improved. ![](https://www.fof-leagues.com/WilleB/Glenn34.png)
|
|
|
Post by rush27 on Sept 27, 2018 19:01:36 GMT -5
My head hurts
|
|
|
Post by Nick on Sept 28, 2018 1:41:56 GMT -5
I just had to go in and fix it again...isn't very stable as it's so easy to change I ran stafford through per screenshot above and get +3.3 My contention (not strongly held) is it doesn't mean more for this guy to be +3.3 than for a 40/40 to be +3.3. Not sure what that means about the number of picks thrown though. Obviously rather have the higher rated guy, all else constant That seems about right as staring at him just guessing i would think his bar is a 50 maybe 60 rated avoid int. On a fundamental level what we are doing is is showing if the players AI is over or under HIS overall rating. Stafford is a 90 rated guy and his AI came out over that figure, +3.3. So we know his AI is over 90. It is a personal calculation for each player. The figure is relevant to him based on his overall rating. So 3.3 means something different according to a players overall - his starting point. It isn't possible for him to be, for example a +7. That is only possible for lower rated players. Every players starting point is his overall rating. The lower rated the player the higher the + figure possible.
|
|
|
Post by rush27 on Sept 28, 2018 4:46:16 GMT -5
Devante Arnold came out at a +4.6. Pretty good
|
|
|
Post by Nick on Sept 28, 2018 14:17:36 GMT -5
Z + 5.7 Mariota -3.0 Rakkoon +4.0 Perry +1.4 Garciaparra +0.8
The Bum +7.9 :-) Huntely -14.1 :-(
|
|
|
Post by thefatcat on Sept 28, 2018 14:23:23 GMT -5
I guess i need to find Feigenbaum a nickname b4 "the bum " sticks!
|
|
|
Post by Nick on Sept 28, 2018 14:28:00 GMT -5
Z + 5.7 Mariota -3.0 Rakkoon +4.0 Perry +1.4 Garciaparra +0.8 The Bum +7.9 :-) Huntley -14.1 :-( So, Huntely is 60 rated. Which means his AI could be a maximum of 60 points less than his overall, hence the room for a huge 14.1 rating. The Bum is 65 overall. Only 35 points possible increase for him to a max of 100 AI. So now just musing...if we took 1.0 to equal 4 points of AI. The Bum would be 7.9 x 4 = 97 AI rating. Huntely would be 14.1 x 4 = 4 AI rating. Z = 74 AI Mariota =53 AI Rakkoon = 83 AI Perry = 87 AI Garciaparra = 49 AI Further thoughts to come on this. But this calculation would seem to be a possible fit. Not convinced it is right, but I think it could be pretty close to the truth. If we could find player better or worse than the Bum and Huntely, we may be able to further refine the 4.0 I used.
|
|
|
Post by Joel on Sept 28, 2018 14:53:48 GMT -5
Z + 5.7 Mariota -3.0 Rakkoon +4.0 Perry +1.4 Garciaparra +0.8 The Bum +7.9 :-) Huntley -14.1 :-( So, Huntely is 60 rated. Which means his AI could be a maximum of 60 points less than his overall, hence the room for a huge 14.1 rating. The Bum is 65 overall. Only 35 points possible increase for him to a max of 100 AI. So now just musing...if we took 1.0 to equal 4 points of AI. The Bum would be 7.9 x 4 = 97 AI rating. Huntely would be 14.1 x 4 = 4 AI rating. Z = 74 AI Mariota =53 AI Rakkoon = 83 AI Perry = 87 AI Garciaparra = 49 AI Further thoughts to come on this. But this calculation would seem to be a possible fit. Not convinced it is right, but I think it could be pretty close to the truth. If we could find player better or worse than the Bum and Huntely, we may be able to further refine the 4.0 I used. I like it. Even if it’s not 100%, it’s a fit that can be used for all
|
|
|
Post by Nick on Sept 28, 2018 15:11:18 GMT -5
I think 4.0 could be a blink too high. If we can get a player under 0 or above 100 using 4.0 we know it is.
|
|
|
Post by Nick on Sept 28, 2018 15:20:03 GMT -5
So yeah Jeremy's Stafford breaks the model. Forget 4.0. Use 3.33*
|
|
|
Post by Strubbel94 on Sept 28, 2018 16:35:41 GMT -5
Wisniewski came back as -0.9. However he just through 14 picks in his last 28 games.
TD/INt ratio per game in those last 28 starts is 2.18/0.5
So im not sure about that to be honest
Edit:
Wisniewski 67 overall.
-0.9 * 3,33 = -2.997
67 + (-2.997) = 64 AI
Is that our best guess ? I think 64 is okay if that calculation is near the truth
|
|
|
Post by thefatcat on Sept 28, 2018 17:00:13 GMT -5
You have to remember that AI is a base rating. Game and player factors will change that rating a lot
|
|
|
Post by Nick on Sept 29, 2018 3:45:45 GMT -5
Great work Squirrel I think this is just about cracked now. I will run more guys on both extremes through it. Updated 3.33* to 3.0 after further analysis.
Updated AI's
Jeremy's pretend QB Stafford 99 AI
The Bum would be 7.9 x 3.00 = 89 AI rating. Huntely would be 14.1 x 3.00 = 18 AI rating.
Z = 68 AI fully developed, 30 AI right now. Mariota = 56 AI Rakkoon = 79 AI Perry = 86 AI Garciaparra = 49 AI Wiesnewski = 64 AI Garrison = 70 AI Antonik = 86 AI (Fulton = 61 AI) Arnold = 65 AI Greathouse = 88 AI Delcorio = 65 AI Hutchings = 79 AI
I think all the of the guys who visually look to have great AI work. Now looking around the league for the worst AI guy I can find. Should be right around 0 AI....
Random bum who visually appears to have awful AI - Colts QB Erikson. Uh, he comes in at -9.4 x 3.3 = 31.02. This brings him in at -3 AI, which isn't possible. To come in at zero we have to bring 3.33* down to 3. So 3.00 flat appears the figure at this point. Although I will continue to see if I can find any guys which force me to lower this figure any further. All results above are updated. Really makes next to no difference.
Another horrible looking AI guy Manke (Packers) = 7.0 x 3.0 = 21. = 9 AI.
|
|
|
Post by Bighouserulez on Sept 29, 2018 12:33:04 GMT -5
OK so now i have been working on a sheet that breaks down Sole/SR/read defense to set a bar for the AI bar. We know the combines can lie but it gives us a step in the right direction.
|
|