Wonky live updates as usual

[chazzycat]

Originally Posted by Drivehud.com

I'm going to assume everyone playing this game understands sample size and variance in baseball.

also, you are giving people WAY too much credit here

[BatAround]

Originally Posted by chazzycat

also, you are giving people WAY too much credit here

As proven by the same people starting threads and posting in threads about the ratings, and then the devs have to come in and explain...again...how the ratings are computed. If you think the algorithm should be adjusted faster/slower that's one thing. To complain about performances that aren't factored in the middle of the year is lazy and just being a complainer.

[Drivehud.com]

Originally Posted by chazzycat

Sure, why not just keep giving players +8 points every month regardless of performance. Do you truly think hitting .260 for a month is worthy of an 8 point bump for that month?? I'm having a hard time with that one. Ohtani I'm not going to argue because I honestly don't know how it works for players that are completely unprecedented in modern times.

That's clearly not what I'm saying.

We're 2/3rds of way through the season. For the live cards to have any relevancy, I'd like to be playing and collecting cards that are more reflective of the season that player is having. There's too much weight on their historical performance imho. Players can have one amazing season and 5 crappy seasons. I think it would be more fun if the algorithm used to rate players reflected that.

[chazzycat]

Originally Posted by Drivehud.com

That's clearly not what I'm saying.

We're 2/3rds of way through the season. For the live cards to have any relevancy, I'd like to be playing and collecting cards that are more reflective of the season that player is having. There's too much weight on their historical performance imho. Players can have one amazing season and 5 crappy seasons. I think it would be more fun if the algorithm used to rate players reflected that.

Ok, sorry for mis-characterizing your remarks. But really, they ARE moving quite a lot, over 20 points/2 tiers each. And like you said the season is only 2/3rds of the way done. They still have time to make it to perfect. I just don't see why folks would be so upset, other than losing money on spec. Not overreacting to hyped players and having a solid statistical approach to everything is exactly the kind of game I want to be playing.

[frankbama]

Originally Posted by chazzycat

Ok, sorry for mis-characterizing your remarks. But really, they ARE moving quite a lot, over 20 points/2 tiers each. And like you said the season is only 2/3rds of the way done. They still have time to make it to perfect. I just don't see why folks would be so upset, other than losing money on spec. Not overreacting to hyped players and having a solid statistical approach to everything is exactly the kind of game I want to be playing.

Interesting one says "only 2/3" and drivehud is saying it "already 2/3". Definitely relevant. Both mean something important here.



Things have changed this year with monthly updates. I'd like to see every 2 weeks or something - monthly is too long and weekly is too quick. This definitely affects expectations. Should a whole month's sample affect the ratings like it was 4 weeks worth from last year or literally looking at that month's end?



Some of the explanations have been different which is why it can get confusing - we heard before that there is more weight given to present year's stats as we go further into the year. This would seem to indicate that their present stats and their last month's relative to the previous month are now more important then their original projections (which I think many people thought weren't so great). This doesn't seem to be the case in a lot of cases.



A lower level example but to get off the diamond/perfect train for a second, look at Keegan Akin. He's been terrible all year, terrible over the last month and ended the month on the Covid IL. Why would his ratings go up? Even in the smaller size his walk and hr rate were roughly the same. Struck out 11 in 10 2/3 vs. 21 in 23. is that tiny k improvement really worth the total rating boost to bronze? Matt Harvey has a had a few starts' success recently but has been terrible all year and it seems like if he made a big jump because of 2-4 recent starts that seems off too. I can't believe that if this was close to the projection for the year, then the projection is not worth a bronze to begin with. Just trying to throw out both ends of the spectrum here.


I feel like people might agree that the real-life logic and the technical logic in the programming need to be brought closer together. I forgot what the actual terms were but in my decades-ago programming classes it was always about making sure the output was correct, but also that it made logical sense - meaning the inputs were correct. Data might mean more than the eye test, but if the data is wrong or interpreted incorrectly than the output doesn't meet the logic test.


Writing too much, but the confusion is well, confusing. Your example of Yelich - doesn't Lemahieu have an extensive track record? Maybe not an MVP but he actually is a low silver at this point.

[Drivehud.com]

Originally Posted by chazzycat

Ok, sorry for mis-characterizing your remarks. But really, they ARE moving quite a lot, over 20 points/2 tiers each. And like you said the season is only 2/3rds of the way done. They still have time to make it to perfect. I just don't see why folks would be so upset, other than losing money on spec. Not overreacting to hyped players and having a solid statistical approach to everything is exactly the kind of game I want to be playing.

People are all-stars after just over half a season. Baseball recognizes players who are having breakout seasons, even if they never have before, why can't OOTPB? There's easy ways to create weighted algorithms that add more weight to a players ranking the further we get into the season. If a player has an OPS over 1k nearly 2/3rds into the season, that player should be rated perfect, and we shouldn't have to wait until the season is over for that to be reflecting in the ratings.

This has nothing to do with over reacting to hyped players, and everything to do w/ how the developers want to define what a live card actually is. Is it a card that puts 90% emphasis on past season performances, or one that uses that data, but also recognizes break out amazballs seasons.

To me right now, the whole live card thing is dumb and useless on so many levels, to the point I'd separate out leagues to only live and only historical if I were the developers. It just makes no sense at all how it's setup now.

[Play4Thrills]

Originally Posted by chazzycat

Ohtani I'm not going to argue because I honestly don't know how it works for players that are completely unprecedented in modern times.

This is the only complaint I had. I followed a lot of other players ahead of this update, and have no other real complaints, because I could see both sides of it -- excepting a few defensive standouts that are truly underappreciated, but I now understand the reasons behind this so I'm not complaining.

The issue I have with Ohtani's low diamond rating is that he is unequivocally the best player in baseball at this point in the season. If we're playing a pickup game of baseball with all living players available, and you get to pick first out of all the other teams, who's on your team playing that day? 100% everyone with any sense picks Ohtani first (assuming no character attachment to another player).

So, it goes to what others have said above -- are Live cards intended to reflect the current value of MLB players? I think that's the idea, and I think we'd all be a little disappointed if we didn't think we could click "Live Start" on a new game and see how teams might finish the season. And if that is the idea, then it's not working, because even after allowing for a modest pace of changes due to sample sizes and a difference of opinion on what you can expect going forward from a player, it's impossible for Ohtani not to be considered the best player in baseball, let alone one of the top 10 or so who make perfect.

Anyway, I don't like to complain. I'm happy that the changes exist and I look forward to the next batch.

[chazzycat]

Originally Posted by Drivehud.com

People are all-stars after just over half a season. Baseball recognizes players who are having breakout seasons, even if they never have before, why can't OOTPB?

All stars are highly non-data driven and shouldn't be used in favor of arguing anything.

Also, you keep ignoring the fact that these players have risen from silver to diamond level. How is that not recognizing a breakout season? It is, just not enough for your taste.

Originally Posted by Drivehud.com

There's easy ways to create weighted algorithms that add more weight to a players ranking the further we get into the season. If a player has an OPS over 1k nearly 2/3rds into the season, that player should be rated perfect

I thought you said people who play this game understand sample sizes? While some statistics can be useful in 2/3rds of a season, like strikeouts and walks are fairly believable...there are others that aren't. BABIP would be the most notorious stat that takes a long time to stabilize, multiple seasons. Batting average being such an important component of baseball, this is a big part of why projection systems put a lot of weight on the last couple seasons, and why OOTP is right to do so. Here is some really good information on sample sizes: https://library.fangraphs.com/principles/sample-size/

To use Vlad as an example again, he was putting up a BABIP of .350 prior to this latest update, on the strength of a 21% line drive rate. Line drive rate is another thing that takes a lot of data to stabilize (600 balls in play per my link above, while Vlad has about 250 this season)

What happened this last update was his line drive rate dipped down to 15%, and with that, his BABIP also dropped 100 points to .250. These are exactly the types of things that projection systems and OOTP are trying to account for, by regressing things to the mean, because there just wasn't enough data to believe that 21% line drive rate yet. And why if he gets back on track with spraying more line drives, the system will believe it more and more as he gets closer to the stabilization point.

Originally Posted by Drivehud.com

This has nothing to do with over reacting to hyped players, and everything to do w/ how the developers want to define what a live card actually is. Is it a card that puts 90% emphasis on past season performances, or one that uses that data, but also recognizes break out amazballs seasons.

I think it does have a lot to do with hyped players. Why else would people be mad that the two most hyped players in MLB today aren't perfect yet? Because they have bought into the hype and not looked closely at the numbers, in Vlad's case. Again I can't really speak to Ohtani.

Your 90% claim seems way off base considering these players are diamonds up from silver.

Originally Posted by Drivehud.com

To me right now, the whole live card thing is dumb and useless on so many levels, to the point I'd separate out leagues to only live and only historical if I were the developers. It just makes no sense at all how it's setup now.

Fair enough. I disagree, I think live cards are a fun component to the game that lets you profit if you're good at analyzing baseball stats, which is appropriate in a game geared towards baseball stat nerds. But that's a separate issue kinda beaten to death already.

[Drivehud.com]

Originally Posted by Play4Thrills

This is the only complaint I had. I followed a lot of other players ahead of this update, and have no other real complaints, because I could see both sides of it -- excepting a few defensive standouts that are truly underappreciated, but I now understand the reasons behind this so I'm not complaining.

The issue I have with Ohtani's low diamond rating is that he is unequivocally the best player in baseball at this point in the season. If we're playing a pickup game of baseball with all living players available, and you get to pick first out of all the other teams, who's on your team playing that day? 100% everyone with any sense picks Ohtani first (assuming no character attachment to another player).

So, it goes to what others have said above -- are Live cards intended to reflect the current value of MLB players? I think that's the idea, and I think we'd all be a little disappointed if we didn't think we could click "Live Start" on a new game and see how teams might finish the season. And if that is the idea, then it's not working, because even after allowing for a modest pace of changes due to sample sizes and a difference of opinion on what you can expect going forward from a player, it's impossible for Ohtani not to be considered the best player in baseball, let alone one of the top 10 or so who make perfect.

Anyway, I don't like to complain. I'm happy that the changes exist and I look forward to the next batch.

You're complaining too much man! I mean Nathan Evoldi is +26 and 98 rated, and Shoei is the best player in the game, +23 and 94.

The ratings make total sense to me.

[Drivehud.com]

Originally Posted by chazzycat

All stars are highly non-data driven and shouldn't be used in favor of arguing anything.

Also, you keep ignoring the fact that these players have risen from silver to diamond level. How is that not recognizing a breakout season? It is, just not enough for your taste.

I thought you said people who play this game understand sample sizes? While some statistics can be useful in 2/3rds of a season, like strikeouts and walks are fairly believable...there are others that aren't. BABIP would be the most notorious stat that takes a long time to stabilize, multiple seasons. Batting average being such an important component of baseball, this is a big part of why projection systems put a lot of weight on the last couple seasons, and why OOTP is right to do so. Here is some really good information on sample sizes: https://library.fangraphs.com/principles/sample-size/

To use Vlad as an example again, he was putting up a BABIP of .350 prior to this latest update, on the strength of a 21% line drive rate. Line drive rate is another thing that takes a lot of data to stabilize (600 balls in play per my link above, while Vlad has about 250 this season)

What happened this last update was his line drive rate dipped down to 15%, and with that, his BABIP also dropped 100 points to .250. These are exactly the types of things that projection systems and OOTP are trying to account for, by regressing things to the mean, because there just wasn't enough data to believe that 21% line drive rate yet. And why if he gets back on track with spraying more line drives, the system will believe it more and more as he gets closer to the stabilization point.


I think it does have a lot to do with hyped players. Why else would people be mad that the two most hyped players in MLB today aren't perfect yet? Because they have bought into the hype and not looked closely at the numbers, in Vlad's case. Again I can't really speak to Ohtani.

Your 90% claim seems way off base considering these players are diamonds up from silver.

Fair enough. I disagree, I think live cards are a fun component to the game that lets you profit if you're good at analyzing baseball stats, which is appropriate in a game geared towards baseball stat nerds. But that's a separate issue kinda beaten to death already.

Vlad jr isn't overhyped. Line drive% actually takes way less to normalize than if his gb/fo ratio was higher, and his BABIP was still high. Then you'd start to question the variance of his primary stat lines. But it's not.. .and here's why...

Line drive% COMBINED with average exit velocity is what matters most, and I think jr is 3rd in the league in that respect right now. To put it simply, the more line drives and the harder they are hit, the higher BA will be. We can take that data and combine it with his average launch angle on fast balls, and see how far out of whack his HR numbers are, which are actually UNDER that the expectation is in this case.

I'm not going to argue with you because I can see it's pointless. You're not even analyzing this data correctly you supposedly are using to justify this rating system. You enjoy it, good for you. It's not logical, and it could be way better if more thought was put into it. But it's a game and I've come to accept it as such.

[chazzycat]

Originally Posted by Drivehud.com

Line drive% COMBINED with average exit velocity is what matters most, and I think jr is 3rd in the league in that respect right now. To put it simply, the more line drives and the harder they are hit, the higher BA will be. We can take that data and combine it with his average launch angle on fast balls, and see how far out of whack his HR numbers are, which are actually UNDER that the expectation is in this case.

wat? Please explain how this fits with his statcast data (xWOBA, xSLG, etc.) all coming in lower than his actual numbers?

You're right, there is data for that. It shows the complete opposite of what you suggest. He's outperforming his expected slugging by 60 points!

[Drivehud.com]

Originally Posted by chazzycat

wat? Please explain how this fits with his statcast data (xWOBA, xSLG, etc.) all coming in lower than his actual numbers?

You're right, there is data for that. It shows the complete opposite of what you suggest. He's outperforming his expected slugging by 60 points!

Maybe because it's not? You don't know what you're talking about, and like I said, I'm done.

https://baseballsavant.mlb.com/leade...ted_statistics

When you can explain Evolidi's ratings over Ohtaini's and use some actual logic, I'll listen. Until then, I'm out.

[chazzycat]

so you're saying .451 is not higher than .428, and .661 is not higher than .606. Got it

[Drivehud.com]

Originally Posted by chazzycat

so you're saying .451 is not higher than .428, and .661 is not higher than .606. Got it

Do you have stake in this company or something?

What I'm saying dear chazz, is that even with those differences, he's what... 4th in the league in xwOBA correct? .001% behind Ohtani. So I stand corrected, I said he's 3rd... he's actually 4th in the league. So even while he's beating expectation, he's still top 5 in the league and should be a perfect at this point in the season if this wasn't a toy game, but it is.

What an overhyped piece of AA garbage. lol

[chazzycat]

Nice job moving the goalposts. We were just talking about him over/underperforming his statcast data, but now we're talking how he performs in MLB rankings of exit velocity. Cool cool.

But I'll bite anyway because I think I can still make my point. See that guy at #1 in xWOBA? His name is Aaron Judge, and he's never had a perfect card. Who's #2...Max Muncy. When was he perfect? Oh that's right, never. It's almost like, hitting the ball hard does not guarantee a card reaches perfect. The guys on that leaderboard who do have perfect cards, also play defense. Vlad does not have much defensive value.

And never did I say anything close to him being "AA garbage", so thanks for the straw man (you're on a roll with the disingenuous arguments!).

I actually think he's amazing! And having a diamond card at age 22 is befittingly impressive. But I can still understand not having a perfect card, and generally agree with their cautious approach to these things.

[Drivehud.com]

Originally Posted by chazzycat

Nice job moving the goalposts. We were just talking about him over/underperforming his statcast data, but now we're talking how he performs in MLB rankings of exit velocity. Cool cool.

But I'll bite anyway because I think I can still make my point. See that guy at #1 in xWOBA? His name is Aaron Judge, and he's never had a perfect card. Who's #2...Max Muncy. When was he perfect? Oh that's right, never. It's almost like, hitting the ball hard does not guarantee a card reaches perfect. The guys on that leaderboard who do have perfect cards, also play defense. Vlad does not have much defensive value.

And never did I say anything close to him being "AA garbage", so thanks for the straw man (you're on a roll with the disingenuous arguments!).

I actually think he's amazing! And having a diamond card at age 22 is befittingly impressive. But I can still understand not having a perfect card, and generally agree with their cautious approach to these things.

Chazz, you're more interested in winning an argument than having a genuine discussion, that's why I made that last comment, because you've been hyperbolic this whole time. It's pretty obvious you're someone who will argue until the cows come home so even though I failed to exit earlier, I will now.

I didn't move the goalpost, I've had the same argument this entire time. Your defense was that he was over performing his stats, which is true. And he already moved up the ranking ladder, but because he's over performing, shouldn't be a perfect card. But as I pointed out, he's still in an essential tie for 3rd as the best overall hitter by expectation in the league, even though he's over performing. Your point is moot.

Have a nice day.

[chazzycat]

Originally Posted by Drivehud.com

I didn't move the goalpost, I've had the same argument this entire time. Your defense was that he was over performing his stats, which is true. And he already moved up the ranking ladder, but because he's over performing, shouldn't be a perfect card. But as I pointed out, he's still in an essential tie for 3rd as the best overall hitter by expectation in the league, even though he's over performing. Your point is moot.

Have a nice day.

You absolutely did, and now you're trying to put words in my mouth. My last few posts have been about rebutting your exit velocity nonsense...I would never have gotten into that, if you hadn't brought it up. On that front, if you would like to have a productive conversation like you say, it really helps to be informed. Your idea that exit velocity + line drive rate is important, simply is not true. Here is a good source to learn how line drives, exit velocity and BABIP actually interplay: https://www.fantraxhq.com/sabermetri...contact-babip/

Hitting line drives consistently is just, far more important than hitting them super hard. So when a line drive rate dips significantly, BABIP tends to also. And that's exactly what happened this latest update with Vlad. And exactly why you need a lot of data to judge BABIP with confidence.

That being said, all of this analysis is completely unnecessary. My original point, before this got so convoluted, was very simple. His performance dropped off by 100 points from his slash line this update, so he didn't keep gaining points.

You have a nice day as well.

[frankbama]

Ok fellas let's move on.



A question - is there an upper limit then? I don't ever think Mullins or Peralta or other randoms on a certain level will be perfect, but what's their upper limit in general terms if they stay doing what they're doing? I don't mean to single those two out necessarily but players like them - gold? Or if we're going with prior history, what's the upper limit on whomever wins ROY?

[bailey]

Pitchers are easy to predict because it's just BB, K, and HR which is all neatly summarized by FIP. Batters has have 5 things to compute: contact (BABIP), gap (XBH), power (HR), Eye (BB), AvK (K). No single number like WOBA summarizes all of those. I'm guessing in Vlad's case that 1 or more of those 5 things were below projections during the period and that is why his OVR did not go up even those his WOBA was 40 points more than projection. On the other hand, last update period there was Buxton who was injured for most of the period and had like 12 PA, but still climbed to 100 and no one complained about THAT wonkiness.

[frankbama]

Can't complain about everyone I guess. FIP is definitely helpful as is looking at each rate individually though it seems RP really got screwed this time.