PING: Everyone. Group Project to study Booms/Busts

[Ben E Lou]

I have a couple more theories regarding booms and busts that I'd like to test out, but in order to do so, I'm going to need more data than I have available to me from just my own careers. Specifically, I'd like to study combine ratings of players who fit one of the two following criteria:

1. Over time, his current rating ended up being 20 points or more above his original potential rating. (boom players)
2. His current rating ended up being 20 points or more below his original potential rating and it was not due to an injury. (bust players)

In other words, for this study I'm defining a boom player as a guy who was 21/38 before the draft, but ended up 58/58, 4-8 years down the line, and a bust player as a guy who was 21/60 before the draft, but ended up being 40/40 a few years later.

I've started this study using my own careers, but just don't have enough. For example, I only have 9 boom QB's to look at, 2 RB's, etc. etc. etc. I'd like to have at least 15 players at each position to examine. (Don't worry about kickers and punters.)

In order to get this study and report findings to y'all, here's what I need:

1. People who do savegames like this (or are willing to start doing it this way): "Falcons 2004 Predraft", "Falcons 2004 Yearend", "Falcons 2005 Predraft", "Falcons 2005 Yearend", etc.. In other words, I want to know the player's combine and scouted ratings BEFORE he was drafted.

2. I want to track the following predraft ratings:

Volatility
Solecismic Test
40-Yard Dash
Bench Press
Agility
Bureau Rating (non-adjusted)
Current Value
Future Potential

3. Here is a suggested format for the data. If you have excel and use the WYSIWYG editor, you can create a spreadsheet like the one below, and copy and paste it directly into this thread.

QUARTERBACKS
Year 1	Name	Cur	Fut	Vol	Sol	40	BP	Agi	Bur	Dev
2019	Anthony Ashmore	18	45	24	34	4.81	19	7.28	3.7	40%
2019	Lawrence Schreiner	17	48	76	35	4.77	15	7.43	4.0	35%
2019	Jorge McGee	17	51	82	37	4.76	11	7.69	4.2	33%
2020	Andre Tanner	19	44	85	39	4.81	10	7.96	3.7	43%
2024	Lawrence Uno	10	44	90	40	4.60	14	7.20	3.6	23%
2026	Neal Ishler	20	50	48	38	4.91	7	7.44	4.0	40%
2026	Shawn McKenzie	16	50	38	39	4.54	14	7.61	4.1	32%
2027	Rob Wiese	7	52	14	56	4.84	14	7.20	4.2	13%
2027	Russell Winters	23	45	64	31	4.54	14	7.60	3.8	51%
RUNNING BACKS
Year 1	Name	Cur	Fut	Vol	Sol	40	BP	Agi	Bur	Dev
2020	Johnny Price	29	38	87	20	4.52	17	7.46	3.5	76%
2026	J.R. Woods	28	33	83	20	4.51	15	7.49	3.1	85%

4. I don't think I need to know the max of their current ratings. I just need to know that the only players we're looking at varied by 20 or more from their original projection.

5. I definitely want these guys divided by positions. My working theory is this: No combine ratings can be ignored, but some are more important for certain position groups. As you can see in the list of breakout QB's above, virtually all of them have the following in common:

1. Just about every single combine rating is better than the position average.
2. Current and future ratings greater than one standard deviation from the position average.
3. Most had above average volatility.

The three factors above were discussed at length in this thread as being important in identifying which rookies to draft. However, I *think* there might be one more factor: look at the Solecismic test scores. The lowest test score of these breakout QB's was 31, which is consider "High" overall, and above the QB position average. Further, seven of these nine guys had Sol. test scores that were more than one standard deviation away from the position average of 28.02. I'm speculating that the fourth important factor for QB's is the Solecismic test in particular, but I'd like to have 5-10 more QB's to study to test this theory out.

Next, take a look again at the two RB's:

Year 1	Name	Cur	Fut	Vol	Sol	40	BP	Agi	Bur	Dev
2020	Johnny Price	29	38	87	20	4.52	17	7.46	3.5	76%
2026	J.R. Woods	28	33	83	20	4.51	15	7.49	3.1	85%

With only two to look at, it is obviously much hard to draw conclusions, or even speculate but these two things stand out to me.

1. Solecismic test might not be as important for them. Both of these guys actually came in BELOW the position average of 21.29.
2. Both of them had off the charts 40 time's for RB's--just about 2 standard deviations from the average.

Early speculation would indicate that 40 time is the most important combine score for RB's, but obviously we need more data to draw such a conclusion.

So, if those of you with mature careers who do two saves per season can start looking through your HOF's, QB Comparisons, Career Leaders, and Top 20 Current Value players by positions and finding those late-round picks and undrafted FA's who wound up being stars, and post their pre-draft info, and then also look through the list of former #1's and FA's on the scrap heap to find those first-rounders who never amounted to much, and post their pre-draft info as well, perhaps we can make some more headway on this.

NOTE: I fully expect that it may take a few weeks to get enough players to examine, so those of you who haven't been saving twice per season who want to help out, start doing so now!

Thanks in advance!

--Ben

[gstelmack]

Just a tidbit: you don't need saved games. FOF Extractor can dump your combine ratings (and blue/gray bars) for the entire draft class to a spreadsheet.

[cthomer5000]

I wished I'd saved my games like that. I have one career in 2060, but I have no access to combine ratings.

[Ben E Lou]

Originally Posted by gstelmack

Just a tidbit: you don't need saved games. FOF Extractor can dump your combine ratings (and blue/gray bars) for the entire draft class to a spreadsheet.

Ahhh! Didn't think of doing that. You can then just save a spreadsheet each season with the entire draft class, huh?

[MJ4H]

I volunteer to study busts.


NOTE: I don't even own any FOF game beyond the original and TCY.

[Barkeep49]

I would love to help, but would like to know if there is an "easy" way to find these booms and busts? In other words how do you track so many players?

[Ben E Lou]

Originally Posted by Barkeep49

I would love to help, but would like to know if there is an "easy" way to find these booms and busts? In other words how do you track so many players?

Booms are easy to find. Several places to look for them:

1. Individual Statistics
2. Select a position by "Current Rosters" and hit Recommend. You'll get the top 20 players at that position.
3. Career Leaders
4. Hall of Fame

Just go to any of those screens, and use the "N" and "P" hotkeys to cycle through the list quickly. Just look for guys drafted in the 4th round or later among the "elite" lists above. Most/all of them will be guys who increased significantly from their pre-draft numbers. The bigger issue is actually having the predraft numbers on hand then to see what they were. That is where having an extra savegame per season, or saving each draft class with the FOF Extractor is needed.



Busts are a bit more tedious. There are two ways that are relatively painless:

1. Track guys on your own team who bust over time.
2. Look at the whole FA list during the season (from the Roster Screen). Cycle through it quickly, looking for former first or second round draft picks who are sitting in the pool.

The additional problem with 1st/2nd rounders with low ratings is that you also have to check to see if they had a serious injury at some point in their career. This can sometimes just be done by inspection immediately. If they played 16 games for 4 straight seasons, then played 2, and sucked from then on out, you can safely assume it wasn't a bust, but an injury. Other times it isn't as obvious. Hope that helps.

[3ric]

Originally Posted by gstelmack

Just a tidbit: you don't need saved games. FOF Extractor can dump your combine ratings (and blue/gray bars) for the entire draft class to a spreadsheet.

I'd love to use FOF Extractor for this, but I can't get it working with my copy of FOF2k4... did you get my email?

[RPI-Fan]

SD: Shouldn't we be making a note in the spreadsheet whether said player is "Bust" or "Breakout". FWIW, I've been running a solo "career" JUST in preparation for the draft of the multiplayer league I'm in. Thus, I'm flying through seasons and just seeing how well I can draft. Doing 3 saves per season - PreDraft, PostTraining, PostSea.

Also, if people start new careers to study this and aren't playing a "real" game in these careers, turn injuries to 0. Will make busts MUCH easier to identify.

[Ben E Lou]

Originally Posted by RPI-Fan

SD: Shouldn't we be making a note in the spreadsheet whether said player is "Bust" or "Breakout".

1. Yeah. Actually I was hoping/assuming that people would make two different spreadsheets: one for each.

FWIW, I've been running a solo "career" JUST in preparation for the draft of the multiplayer league I'm in. Thus, I'm flying through seasons and just seeing how well I can draft. Doing 3 saves per season - PreDraft, PostTraining, PostSea.

I think that's a very smart move.

Also, if people start new careers to study this and aren't playing a "real" game in these careers, turn injuries to 0. Will make busts MUCH easier to identify.

Duh. Good call. I was assuming I'd be getting data from existing careers, because several people have mentioned that they do the predraft and postsea saves as a matter of course, but it didn't occur to me that some might be able to just do some drafting for the purpose of testing this.


FWIW, right now my working theory is that while breakout players have some level of predictability, bust players are much more random. If this is true, I'd say it is a good thing.

[Ben E Lou]

Bump

[Barkeep49]

I've generated a bunch of seasons, just haven't had time to pull the people. Will try and devote some time to this soon.

[3ric]

Would a condition for the validity of the study be that only in-game rosters and draft files are used? If you used HeavyReign's roster generator and TCY draft files, would that affect the outcome significantly?

[Neuqua]

dorks

[stevew]

Is "formations known", another component to QB success?
Just curious, cause I had a breakout QB....went from like 19/43 to 40/64 almost overnight. He did have good combine scores IIRC though.

And do you have the wig draft combine scores greg? Id like to look at a couple breakouts I've already noticed.

[gstelmack]

Originally Posted by stevew

And do you have the wig draft combine scores greg? Id like to look at a couple breakouts I've already noticed.

I'll have to check if I still have that spreadsheet at home. If not, Wig had the spreadsheet up on the web site, he might still have the HTML.

And one of these days I'm going to add "formations known" to the QBs in the Extractor.

[Lucky Jim]

Using HR's utility wouldn't matter because it only affects the starting rosters. Draft players are still generated by the game in the same way. This should be an interesting study, I'll contribute what I can.

[Tium]

Hi,
I've tried to track boom and bust players following the guidelines posted by SD here and I'd like to share the results I got. The analisys is restricted to a short career (10 year span) but it could be interesting enough.
It's not that difficult to find boom players but it's not easy to find busts, just because I can't know if they had any serious injury during their careers. So I've tried to look at their career stats, trying to decide if they have been healthy or not, reporting only the ones fitting this requirement, based on my judgement.
A short legenda:

- Boom is a player with a current Future Rating 14 points or more greater than the Draft Day Future Rating.

- Bust is a player with a current Future Rating 14 points lower or less than the Draft Day Future Rating.

Some notes:

- Boom players that I could check yearly (those on my team) have a post Training Camp//contract signing Future Rating greater than or equal the same rating showed in the Draft Board. This value goes up every year, steeper the first years.

- Bust players that I could check yearly (again, those on my team) usually drop their rating after the first TC/contract signing and it never goes up later, well... it often goes down

- I can't find a consistent link from the combine values and the boom/bust players. It's true that most great players have at least higher than average values, but you can see some bad players with great combine rating...
I think that could be useful to restrict the boom/bust range to +- 10, so that we could have a larger statistical base. To be honest I'm not satisfied with this small research but it's really boring to find bust players... I hope to post more data in the future.

- It looks like boom and bust are pretty random whilst players with good combine ratings are more likely match the scouted future rating. This is what I'd like to prove with further test.





And

P.S.
Could someone let me know how to paste those Excel cells in the post??? The WISIWYG editor I use doesn't work well Thanks!

EDIT: Tried to make the post looking better...

EDIT2: Images seem to work better...

[Cringer]

it took me 10 minutes to scroll to the end of the Tium post, and not a thing about a trout, oh well.............

[stevew]

It would be a great post if not for the 60+ pages of nothingness

[WSUCougar]

Originally Posted by stevew

It would be a great post if not for the 60+ pages of nothingness

You mean you really can't see anything?

*points finger and laughs derisively*

[stevew]

Originally Posted by WSUCougar

You mean you really can't see anything?

*points finger and laughs derisively*

No, I can see stuff, then it dissolves into 60 pages of nothing.

[Tium]

I do apologize for my poor posting and I'm going to use images until I can find a better way to insert tables...

And