Today's Live Card Update!

[mrbucket]

Ahmed has a high BABIP. His wRC+ is matching his ZiPS projection but because of his BABIP, the system figures it's because he's lucky rather than anything else so he's considered to be underperforming compared to what was projected.

[dkgo]

Originally Posted by herbyhancok

You're misunderstanding me. If the only ratings being accounted for on ratings adjustments are offense *without rating that against the cards base offense values* then the database is just going to say, "Oh, I see this player had a WRC+ below 90 again this week, lemme further dump on this player's offense and overall ratings, despite them having played well in real life."



Hello Nick Ahmed owners.

but it does rate it against the cards base values. I'm sure we can pick a few examples that look funny every week in both directions, but by and large it will effect players equally

my george springer dropped from 84 to 83. offensive-first player who is hitting a very nice .300/.356/.500 this season so who knows what exactly happens at the margins

[Dogberry99]

Originally Posted by herbyhancok

It's not a big deal right now, but if these glove guys are losing, say a point a week on average every update and there's nothing done, then by September most every silver glove guy or catcher is going to be an iron card.

*I don't think the problem is just limited to defense first guys either, it's just there seems to be a more apparent pattern w/ them. Cause if you look at a lot of guys there's a lot of head scratchers. Adam Jones is raking and got a -1.

Adam Jones went up. The DB is wrong as he was an Iron card previously.

I was going to point out that the same thing happened to Isiah Kiner-Falefa, but this type of generalized discrepancy has already been acknowledged.

[herbyhancok]

Originally Posted by mrbucket

Ahmed has a high BABIP. His wRC+ is matching his ZiPS projection but because of his BABIP, the system figures it's because he's lucky rather than anything else so he's considered to be underperforming compared to what was projected.

If it's penalizing players for BABIP I don't like that at all since we're dealing w/ small sample sizes among other things.

Originally Posted by dkgo

but it does rate it against the cards base values. I'm sure we can pick a few examples that look funny every week in both directions, but by and large it will effect players equally

my george springer dropped from 84 to 83. offensive-first player who is hitting a very nice .300/.356/.500 this season so who knows what exactly happens at the margins

Have you even bothered looking at the spreadsheet? 44 batters have received stat bumps for a total of 76 positive, while 204 have received decreasees for a -411 negative total. Far more batters are being penalized and at a higher impact. I'm not cherry picking, this is systemic.

[Charlatan]

Originally Posted by herbyhancok

44 batters have received stat bumps for a total of 76 positive, while 204 have received decreasees for a -411 negative total. Far more batters are being penalized and at a higher impact. I'm not cherry picking, this is systemic.

Didn't I see a post from Marcus saying the league averages were lower than they initially projected? Maybe that's what is behind this big downward adjustment - they're trying to get the league average in OOTP to match the real-life league average. If that's the case then after the big bumps this week the following weeks should see smaller adjustments to players and a positive/negative adjustment ratio of more or less 1:1

[chazzycat]

Originally Posted by herbyhancok

If it's penalizing players for BABIP I don't like that at all since we're dealing w/ small sample sizes among other things.

No one is being "penalized" for having a high or low BABIP for a week. BABIP in this context is just a tool that helps us figure out what's really going on.


Back to Nick Ahmed - his BABIP is .353. That lofty number explains why his overall production has met expectations, in terms of wOBA or wRC+ or whatever stat you like. But if you look closely at his batting line vs. the projections, you will see he is under-performing in both walk rate (Eye rating) and ISO (Power rating). Those are both more skill based than BABIP, which has a huge luck component. So I'm guessing those are the ratings that actually were penalized in the update, not his BABIP rating. BABIP just helps us explain why his production remained high.

[dkgo]

Originally Posted by herbyhancok

Have you even bothered looking at the spreadsheet? 44 batters have received stat bumps for a total of 76 positive, while 204 have received decreasees for a -411 negative total. Far more batters are being penalized and at a higher impact. I'm not cherry picking, this is systemic.

literally explained in the first post

also has nothing to do with your theory that "glove-first" cards are getting hosed

[mpatt75]

Moncada is hitting the crap out of the ball so far. Why didn't his ratings get raised?

Man, y'all forgetting the White Sox exist too?

[herbyhancok]

Originally Posted by chazzycat

No one is being "penalized" for having a high or low BABIP for a week. BABIP in this context is just a tool that helps us figure out what's really going on.


Back to Nick Ahmed - his BABIP is .353. That lofty number explains why his overall production has met expectations, in terms of wOBA or wRC+ or whatever stat you like. But if you look closely at his batting line vs. the projections, you will see he is under-performing in both walk rate (Eye rating) and ISO (Power rating). Those are both more skill based than BABIP, which has a huge luck component. So I'm guessing those are the ratings that actually were penalized in the update, not his BABIP rating. BABIP just helps us explain why his production remained high.

He is literally, being penalized, he has a -3 adjustment. Like I said, small sample sizes, he could just be getting more pitches to hit this week. Let's say one hit, JUST ONE, of his hits was a 3-1 pitch that's outside the zone, which he laced for a single to knock in a runner from 3rd. We can say he lays off that pitch, now he has a .282 BA and a 7% BB rate, both well above his career rates.


You can't penalize guys for increased BABIP like that in such a small sample size. Look at Justin Turner, turns out he wasn't just lucky.

Originally Posted by dkgo

literally explained in the first post

also has nothing to do with your theory that "glove-first" cards are getting hosed

Yeah, I read it. Why can't it be both? This has everything to do w/ what I was saying about glove first guys. Look at the spreadsheet, now look at which position which positions are being penalized the most and at the highest impact. Do you think it's just random variance that one CF got a positive bump and like 30 got a negative bump?

[chazzycat]

Originally Posted by herbyhancok

He is literally, being penalized, he has a -3 adjustment. Like I said, small sample sizes, he could just be getting more pitches to hit this week. Let's say one hit, JUST ONE, of his hits was a 3-1 pitch that's outside the zone, which he laced for a single to knock in a runner from 3rd. We can say he lays off that pitch, now he has a .282 BA and a 7% BB rate, both well above his career rates.


You can't penalize guys for increased BABIP like that in such a small sample size. Look at Justin Turner, turns out he wasn't just lucky.

The point I was trying to make is that he is being penalized for under-performing his eye and power numbers, not the BABIP itself, which is just an indicator.


I guess you could argue that weekly updates are too frequent...but really...a 3 point adjustment does not seem that drastic. I own Ahmed and certainly am not upset about this.

[Dogberry99]

Maximum adjustments of 3 points (assuming this is a max) means that it would be theoretically possible for a player to shift 2 color tiers in less than a month.

Previous comments by the dev team seemed to indicate that shifts of this size would be extremely unlikely or possibly even impossible even when the timeframe is lengthened to the entire year.

[herbyhancok]

Originally Posted by chazzycat

The point I was trying to make is that he is being penalized for under-performing his eye and power numbers, not the BABIP itself, which is just an indicator.


I guess you could argue that weekly updates are too frequent...but really...a 3 point adjustment does not seem that drastic. I own Ahmed and certainly am not upset about this.

If he's being penalized for his eye in such a small sampe size, then he should be getting perks to his contact, which would offset that. Like I said, if you trade just one of those hits for a walk, like in the scenario I drew up, he's well above his norms in average and BB%.


I own Ahmed too; I'm not heartbroken over it by any means, he's just sitting in my collection. I'm not coming to this from a "my guys got dinged and I'm upset" kind of mentality. I probably benefited more than lost on the ratings update. I just noticed a strong trend of many guys getting dinged, probably unfairly - and I feel that there might be something off on how things are being computed and if there is I hope me bringing it up helps it get fixed.

[kyman73]

Turner and Lamb, two gold hitters both go down with injuries last week and will be out minimum 5 weeks; at least 5 full "seasons".

Obviously we should keep their stats frozen then right? Would be unfair to affect the next 5+ "seasons" for their owners based on the 15 & 18 PA's they had over the last week and a half.

In those <20 PA's Lamb OPS'd .722 and Turner was at 1.25 (for reference both players were projected by ESPN right around .775 for the season). Instead of keeping them at their initial rankings your algorithm drops Lamb 3 points and doesn't increase Turner by a similar if not greater amount. In my opinion that's pretty bogus and should be looked into.

[dkgo]

I do think that there are some far too drastic changes based on tiny samples.

Lamb is a good example. FIFTEEN AT-BATS that were slightly below expectation and he gets dropped 3 points?

[UltimateAverageGuy]

Releasing the changes is a REALLY bad idea. Too many people want perfection and their players to get bumped out of selfishness and will clog the heck out of every weekly thread.

[Markus Heinsohn]

Originally Posted by UltimateAverageGuy

Releasing the changes is a REALLY bad idea. Too many people want perfection and their players to get bumped out of selfishness and will clog the heck out of every weekly thread.

I don't worry about this at all, you cannot please everybody.

By the way, we are still working on the process once we get a bigger sample size, so any weird thing that may have happened due to the small sample size will be corrected eventually.

[Markus Heinsohn]

Originally Posted by herbyhancok

The thing I'm worried about is what the ratings changes are based on, like I was asking in my previous post. If defense is static and ratings changes are only based on offense, you basically never want to own glove-first live cards or you're going to get hosed every week. Catchers seem to be way over represented on the list, so this is probably the case. What did Ketel Marte do to deserve this???

This is not how it works. The OVR is still based on defense + offense of course for position players, nothing changes here. The offensive ratings will change with the live updates, and the defensive ratings remain static.

As I said in my initial post, we did change some ratings + some tweaks to the algorithm that may have led to OVR changes for some players that cannot be attributed to the 2019 performance entirely. From now on this will not happen though, so things will progress smoothly.

[herbyhancok]

Originally Posted by Markus Heinsohn

This is not how it works. The OVR is still based on defense + offense of course for position players, nothing changes here. The offensive ratings will change with the live updates, and the defensive ratings remain static.

As I said in my initial post, we did change some ratings + some tweaks to the algorithm that may have led to OVR changes for some players that cannot be attributed to the 2019 performance entirely. From now on this will not happen though, so things will progress smoothly.

I guess I worded that poorly. I was trying to say that if the only dynamic stats affecting the overall rating are offensive, and the weekly stats aren't weighted against the player's expected offensive production, but rather based against league average. That, that would explain why there seems to be a very disproportionate amount of up the middle players with negative ratings adjustments.


Thanks for the response.

[Markus Heinsohn]

Originally Posted by herbyhancok

I guess I worded that poorly. I was trying to say that if the only dynamic stats affecting the overall rating are offensive, and the weekly stats aren't weighted against the player's expected offensive production, but rather based against league average. That, that would explain why there seems to be a very disproportionate amount of up the middle players with negative ratings adjustments.

Performances are weighted towards 2018 league averages, since that is what the original ZiPS ratings are weighted towards, too.

Anyway, we will continue to tweak the process. I'd expect things to be more consistent in next week's update when we have more data to work with

[dkgo]

It does make sense that there was essentially a correction to the original ratings that is showing as "updates" and going forward there won't be so many changes.

I do like the concept, the downside to me is that it creates a lot of "feel-bad" moments for players who didn't really do anything wrong. People are nervous to actually use the best players in the game out of fear that any monday they could wake up and see they lost 15K points in value, and mass quick-selling of fringe perfects/diamonds on sunday night takes them out of circulation in the game economy.

I kept one of my Rendon's. I like using him and he's doing well, but he's not worth 14K points more than say manny machado or matt chapman in terms of actual card quality. so I'm going to have to sell him for one of those guys to make sure I don't lose a bunch of value if he happens to have a bad week.