Using OOTP 16 Simulations of 2015 Season as a Fantasy Baseball Draft Tool

BradWilliamson · 04-03-2015, 11:35 AM

Would you feel comfortable drafting your fantasy baseball team based on your evaluation of the performances that result from OOTP's simulation of the 2015 season?

I want to believe the results would be safe enough to draft by.

But then again... I can't help but be skeptical.

Thoughts?

r0nster · 04-03-2015, 11:49 AM

Yes and no ... would depend on what rating system the drafting being used. Might be much different than OOTP. But stats would most assured follow OOTP

BradWilliamson · 04-03-2015, 11:53 AM

Quote:

Originally Posted by r0nster

Yes and no ... would depend on what rating system the drafting being used. Might be much different than OOTP. But stats would most assured follow OOTP

But what about just using OOTP to get a general idea about whether or not a player's performance will improve or decline in comparison to their 2014 campaign?

Mister_G · 04-03-2015, 03:17 PM

I was thinking about this in preparation for my draft tonight. OOTP uses PECOTA, so I was trying to figure out if there was a way to make the game rank players offensively based on these projections, without having to do multiple sims which will have outliers.

r0nster · 04-03-2015, 06:30 PM

stats help in a major way but just be aware of their rankings and the sites point system and ratings

LordTC · 04-03-2015, 06:53 PM

I'd be skeptical. You can see massive variance in potential and overall for players year over year in OOTP, many of which are predictable: OOTP overfits on players having a single huge year when the data shows those years are outliers and many players are skeptical of repeat performances.

It might be useful to use OOTP to get a series of ratings then look for anomalies or outliers and manually adjust them. It might also be useful to use the progression engine rather than refitting and simulate say 4 different series: 2011,2012,2013,2014 then apply a weighted average -- this would help limit overfitting on single year performance, but you'd have to trust the progression engine which I personally think gives more possible realities than actual realities.

You might want to consider simming with injuries off then manually downgrading injury prone players as well, as the injury engine is quite high variance (you don't want to automatically not draft any pitcher who picks up a major injury in a single run of the sim for example).

04-03-2015, 11:35 AM	#1
BradWilliamson Bat Boy Join Date: Apr 2015 Posts: 6	Using OOTP 16 Simulations of 2015 Season as a Fantasy Baseball Draft Tool *Would you feel comfortable drafting your fantasy baseball team based on your evaluation of the performances that result from OOTP's simulation of the 2015 season?* I want to believe the results would be safe enough to draft by. But then again... I can't help but be skeptical. Thoughts?

04-03-2015, 03:17 PM	#4
Mister_G All Star Reserve Join Date: Jan 2012 Posts: 576	I was thinking about this in preparation for my draft tonight. OOTP uses PECOTA, so I was trying to figure out if there was a way to make the game rank players offensively based on these projections, without having to do multiple sims which will have outliers. __________________ A man doesn't make mistakes, mistakes make the man @MisterG90

04-03-2015, 06:53 PM	#6
LordTC Major Leagues Join Date: Aug 2009 Posts: 479	Fantasy Draft I'd be skeptical. You can see massive variance in potential and overall for players year over year in OOTP, many of which are predictable: OOTP overfits on players having a single huge year when the data shows those years are outliers and many players are skeptical of repeat performances. It might be useful to use OOTP to get a series of ratings then look for anomalies or outliers and manually adjust them. It might also be useful to use the progression engine rather than refitting and simulate say 4 different series: 2011,2012,2013,2014 then apply a weighted average -- this would help limit overfitting on single year performance, but you'd have to trust the progression engine which I personally think gives more possible realities than actual realities. You might want to consider simming with injuries off then manually downgrading injury prone players as well, as the injury engine is quite high variance (you don't want to automatically not draft any pitcher who picks up a major injury in a single run of the sim for example).

04-03-2015, 11:49 AM	#2
r0nster Hall Of Famer Join Date: May 2006 Location: Michigan Posts: 3,081	Yes and no ... would depend on what rating system the drafting being used. Might be much different than OOTP. But stats would most assured follow OOTP

04-03-2015, 06:30 PM	#5
r0nster Hall Of Famer Join Date: May 2006 Location: Michigan Posts: 3,081	stats help in a major way but just be aware of their rankings and the sites point system and ratings