BF.net Brewer Projections Idea

[TerraceReserved]

An idea came to me a while back when looking at some Brewer projections: what if a large group of fans (bf.net) could collectively predict players future performance as well as a complicated prediction system such as Pecota or Zips? There'd be no algorithms, equations, or math at all, just fans geussing how a player will do.

This method would have it's advantages. For example, Brewer nation is primed for a Weeks breakout - something that projection systems probably won't pick up on. Fans can use quantitative and qualitative observations to form the predictions, something projection systems don't do. And with a large enough survey that can be done here at bf.net, there will be pessimistic and optimistic numbers from a wide range of different views.

Predicting pitching would have a big advantage too. As fans we know that the main reason for some of our pitching woes last years was defense, and now our D is improved by leaps and bounds, thus most likely improving our pitchers

Here's the plan: you make a prediction without looking at any projections, but just by looking at a players statistics and your gut feeling on how the player will do to form your own. Any stat is in play to use to form your projection. When all the predictions are in I'd average them up and those would be the official bf.net projection.

My thought was that the only the major stats should be predicted. AVG/OBP/SLG I think would be fine for offense, as predicting every single counting stat would be a royal pain to predict and for me to average. For pitching, just ERA and WHIP, and maybe K's, BB's.

I'm open to suggestion for any changes to this experiment. For example, would it be better to predict everyone on the 25 man, or just the 8 position starters and pitchers? Or only select starting pitchers to narrow it down? Should any more (or different) stats be projected?

[Ennder]

I'd add at least SB to the mix for offensive stats and maybe AB and IP to get a proper framework of playtime.

[rluzinski]

Good idea but if you want people to do better than projections, I don't see why you wouldn't want them to look at the projections (if they want) and combine that with the more subjective information they have. Any information that they want.

 

I'd also restrict the list to maybe 8 position players and 6 pitchers. I'd rather people spend more time on each player than throw out 25 guesses.

[TerraceReserved]

Thanks for the tips End and Russ.

 

Rluz, it does make sense that if they want to projections they can. And 8 position players and 6 pitchers sounds like the best option.

 

I'll be getting my predictions in tomorow, and hopefully there's a good response to this so we can see some meaningful and interesting overall projections.

 

For reference, the stats that should be projected (they're subject to change) are:

Hitting: AVG/OBP/SLG, HR, SB (BB, K's if wanted)

Pitching: ERA, WHIP (K's, BB's, HR's if wanted)

[Ennder]

I'd suggest people project the following guys.

 

Weeks

Cameron

Fielder

Braun

Hart

Hall

Hardy

Kendall

 

Sheets

Bush

Capuano

Villanueva

Suppan

Gallardo

Gagne

Riske

Torres

Turnbow

 

That is more pitchers than hitters but our pitching depth is more interesting than the hitting depth since we have no straight platoon guys this year. You obviously want to project the entire starting rotation but that covers at least 7 guys right there and then at least 2-3 of the bullpen guys are important.

[pogokat]

anyway someone could set up a spread sheet to track this stuff, rather than just having the info appear here? Honestly, this could make a cool article very similar to the fan prospect poll that Patrick just completed...

[rluzinski]

I have no problem with counting stats being included but IP and PA should also be added then, so it can be converted into a rate stat, IMO. For instance, I'd suggest this for pitchers:

 

IP, BB, K, WHIP (BABIP wouldn't work very well), ERA

 

batters:

 

PA, BB, HR, SB, CS, BA, OBP, SLG

 

After terrace is comfortable with the format (it was his idea, afterall), I suggest that he start a new thread with a template for people to fill out (ground rules, players to project and metrics to include).