2021 projections

[Whistler]

Also, I'm noting an oddity on my CDG projected 2021 versus YTD 2020. Perhaps this is just a function of pre-season PT assignments, but for my same team settings in 2021 and 2020, there are 92 players with $values above $15 in 2020, and only 58 players with $ values projected in 2021. I started looking when I noted in my 10 team AL only league that there were only 8 players total (H+P) that earned between the $22.50 and $30 grid lines in 2021 versus 28 in YTD 2020. A few spot checks on PT though doesn't show much to me though. Thoughts?

[Whistler]

A good example of the qualifying games at 20 - Brantley is listed as DH only in CDG - I presume that is because he had 19 games in LF, and the CDG is still set on 20.

Which raises a question. If a player is removed from one column (say DH) and moves to another column (OF) - or if a multi-position player gets less columns -- does that significantly affect their player value, and if so how much? Or just which column there are in or not?

Thanks

[Whistler]

Ah, nice find. You've got much of this right...

- definitely the 2018-19 MLEs are a factor. The Forecaster shows his 2018 BA as 276, 2019 as 268 with MLEs included
- xBA is another big factor: three year xBA 253-260-282
- hit rate for those three years is 31-33-36. We're at 31 in the 2021 projection, that could come up a tick. I'll nudge that to 32 in next projections update, should be good for ~4-5 BA points or so.

In general, this surfaces some of the questions from Ron's intro to the Forecaster about how much weight to put on 2020. Gurriel improved his bb%, ct%, xBA, LD%.... for 224 PA. Is that sustainable? The commentary from BF also raises the question whether he's going to maintain that ct-and-LD approach that's good for BA, or go back to his 2019 upper-cutting ways.

Ah, H%. And small samples for his entire grid as his 2018, 2019, and 2020 are all in the same PA ballpark

With regard to xBA effect, Gurriel is also interesting in that he has good SPD scores, but his SB% and overall SB totals are not. I read the definition of xBA in the Forecaster to include SX, an index that factors in the SB elements (along with his reasonably stable FB/GB/LD and CT%, and PX ), as opposed to the DeCaria pure speed (SPD) that is listed in the Forecaster player boxes. So....does a player with good SPD but bad SX (good speed, bad SB guy) give the player a lower xBA than expected in that his SPD demonstrates he can get to first (or 2nd, or 3rd) faster from the plate (improving his BA and OBP), but his speed once on base (measured by SB and SX) penalizes his xBA when there is no BA or OBP effect once the player is already on base.

Definitely see the problem with error bars in a player like Lourdes without full season MLB track records. He "seems" like an established player that is improving, as he's been around for 3 years and is improving in some metrics over half-seasons. But he only has 206 MLB games under his belt.

[RAY@HQ]

Ah, nice find. You've got much of this right...

- definitely the 2018-19 MLEs are a factor. The Forecaster shows his 2018 BA as 276, 2019 as 268 with MLEs included
- xBA is another big factor: three year xBA 253-260-282
- hit rate for those three years is 31-33-36. We're at 31 in the 2021 projection, that could come up a tick. I'll nudge that to 32 in next projections update, should be good for ~4-5 BA points or so.

In general, this surfaces some of the questions from Ron's intro to the Forecaster about how much weight to put on 2020. Gurriel improved his bb%, ct%, xBA, LD%.... for 224 PA. Is that sustainable? The commentary from BF also raises the question whether he's going to maintain that ct-and-LD approach that's good for BA, or go back to his 2019 upper-cutting ways.

[Whistler]

Trying to wrap my head around the Lourdes Gurriel projections from a technical aspect. His is projected to have .306 OBP and .271 BA, the lowest numbers of his MLB career. I understand regression to the mean, but I'm curious to regress those averages to be career worst

OBP, BA
2018 309, 281
2019 327, 277
2020 348, 308
2021 306, 271


Is this a function of adding in 2018 and 2019 MLE's into the equation?

[RAY@HQ]

Another update: BF21 commentaries now loaded in Playerlink.

[RAY@HQ]

Updating above issues list... ADP is fixed, now showing 2021.

[RAY@HQ]

That was the CDG bullet in my list above. It’s 20 games right now, but we are going to cut that down and/or make it configurable. Our tech team is working on it.

[Red]

One question I have is the number of 2020 games at a position to quallify there in 2021? McNeil had 12 games at 2B and 9 at 3B. The format that I play has reduced the games played required from 2020 abreviated season. This first draft only lists McNeial as a LF and DH. Not a big deal but I imagine that most leagues will reduce the games played requirement this season and allow McNeil at 2B.

[bolton wanderers]

Thanks, Ray and team! Early Christmas.

[RAY@HQ]

Working issue list:

- FIXED BF21 commentaries aren't in Playerlink yet.
- FIXED the last digit in MM scores is wrong... we hacked that for the 60-game season last summer, and now need to un-hack it.
- FIXED ADP data is from 2020, working with NFBC folks to get that feed going again.
- Custom Draft Guide is working, but it's still set for 20g eligibility. We'll at least get this changed to 10, and hopefully add some level of customization a little later in the offseason.
- projections are missing for a whole chunk of lower-PT% players. Right now the most prominent (by PT%) are Yolmer Sanchez and Tyler Flowers on the hitting side, and Joe Ross/Mike Leake on the pitching side (and many more, of course... point being, not a lot of fantasy difference makers). We'll be working through those in descending order of PT% to get them added to the data set.

For now, though, this looks pretty viable. Lots of revisions to come. Enjoy!

[RAY@HQ]

OK, projections are up!

After very quickly knocking around the site, everything appears in order. The projected data looks reasonable. The Custom Draft Guide works. CSV files are available.

The staff and I will do a bunch of data validation on individual players/teams over the weekend, correcting outliers as we go. So don't get too attached to the details yet... but this looks like a plausible starting point.

I'll also post a known issue list in a couple of hours after I've cleaned up the easy stuff.

[Red]

Thank you! Best present I'll get this Chrisitmas! Happy Holidays!

[RAY@HQ]

Yup, I’ve got batters all prepped. Need another hour or so working on the pitchers offline. I’ll post mid-afternoon and then need some time to debug and polish online. Will update here.

[Red]

Still planning to post today? Appreciate your efforts.