Whereas, I have shown multiple players with real stats.
Mike Trout on pace for greatest season in MLB history
- Thread starter marquee moon
- Start date
- Latest activity Latest activity:
Whereas, I have shown multiple players with real stats.
That don’t prove your argument.
deac29
Steve Lepore
- Joined
- Apr 2, 2011
- Messages
- 2,979
- Reaction score
- 414
I should know better, but here goes:
You misunderstand the issue/question. The question is not whether certain batters have better numbers against certain pitchers. Those small sample anomalies clearly exist.
The question is whether those samples are predictive of future results. They are not, particularly compared to a given batters much larger sample of all at bats against all pitchers. Again, this is not opinion, but statistically proven by the research of Tango and others.
And I showed you multiple batters whose previous success against certain pitchers continued over many years. Each had a big enough sample size to be predictive of that batter versus that pitcher.
I've never said that because Player A hits Pitcher B that you can predict Player A will do well against Pitcher C.
ALL I've said (and I've shown) is that if you have a reasonable sample size that says Player A hits better against Pitcher B that the next time it is more predictive that he will do better than his overall average and against other pitchers and will likely continue his success.
Each of the players I showed had over 50 PA against the specific pitcher. As I stated early on 17 is not indicative of anything. One game can skew such a small number.
Again, I will be shocked if you even move 1% off of your position.
The studies posted directly disprove this. You should read them.
Here's another player; Manny Ramirez vs. Roy Halladay
Career - .411/525/.996 1 HR every 14.85 ABs
Vs. Halladay - .286/397/.683 1 HR every 36.5 ABs (73 ABs)
Batting against Halladay those are predictive that he will do worse than against other pitchers/dramatically worse than his career average.
https://www.rotowire.com/baseball/vspitcher.htm?id=3501
You can't predict more accurately than by using the most.
Career - .411/525/.996 1 HR every 14.85 ABs
Vs. Halladay - .286/397/.683 1 HR every 36.5 ABs (73 ABs)
Batting against Halladay those are predictive that he will do worse than against other pitchers/dramatically worse than his career average.
https://www.rotowire.com/baseball/vspitcher.htm?id=3501
You can't predict more accurately than by using the most.
Last edited:
Sorry, I was watching the Preakness and betting.
You can't predict more accurately how someone will do that by using the numbers he has against the opponents he most often faces once they reach a reasonable base.
As to the studies, you start with a terribly low number of ABs to use as a definitive base. I've used career numbers that show among highest usages by the players.
You can't predict more accurately how someone will do that by using the numbers he has against the opponents he most often faces once they reach a reasonable base.
As to the studies, you start with a terribly low number of ABs to use as a definitive base. I've used career numbers that show among highest usages by the players.
Except that the studies say it is more predictable to look at their career performance. Keep being wrong though.
Keep being a condescending prick. You seem to like that persona. I haven't done that to you.
I show you multiple players with positive or negative, consistent out of the norm results and you ignore them. But you are proving what I said early on. I could come up with 100 such players or a 1000 and it won't make any difference. You've already made up your mind and no amount of data that directly disputes your premise will ever be enough.
This is over.
I show you multiple players with positive or negative, consistent out of the norm results and you ignore them. But you are proving what I said early on. I could come up with 100 such players or a 1000 and it won't make any difference. You've already made up your mind and no amount of data that directly disputes your premise will ever be enough.
This is over.
To say that Tommy Hutton wouldn't do better against Tom Seaver than a random pitcher when you have career length stats that prove he did is moronic.
To say that Roy Halladay isn't more likely to get Manny Ramirez out than a random pitcher would when have a career of numbers that say otherwise is irrational.
The difference between a theory (which is what talk about) and the actual results is that one may happen and one did. You can talk about theories all day long, give me actual results.
To say that Roy Halladay isn't more likely to get Manny Ramirez out than a random pitcher would when have a career of numbers that say otherwise is irrational.
The difference between a theory (which is what talk about) and the actual results is that one may happen and one did. You can talk about theories all day long, give me actual results.
RJ, you are still not understanding this point.
You are conflating 2 arguments. I am not saying that player A cannot outperform his baseline statistics against pitcher X. That will certainly happen. And players that have done well will sometimes also continue to do well in a matchup. That doesn't prove anything.
But you are saying that because player A did well over 75 ABs, that it means players who have done well over ~40 ABs, will continue to do well over the next ~40 ABs. That is a separate argument. And you have a selection bias in that you are limiting your data set to the successful.
What I am saying is that if you take 40 players that did above average against a pitcher for 40 ABs, and asked if it was predictive in how they would do in their next ~20 ABs, it is NOT more predictive then just looking at the overall last 1500 ABs of the player. That is proven in multiple studies cited here. You are using the argument that because some players were successful in both the first sample set and the last sample set, it means that the first sample set is always predictive.
But you are saying that because player A did well over 75 ABs, that it means players who have done well over ~40 ABs, will continue to do well over the next ~40 ABs. That is a separate argument. And you have a selection bias in that you are limiting your data set to the successful.
What I am saying is that if you take 40 players that did above average against a pitcher for 40 ABs, and asked if it was predictive in how they would do in their next ~20 ABs, it is NOT more predictive then just looking at the overall last 1500 ABs of the player. That is proven in multiple studies cited here. You are using the argument that because some players were successful in both the first sample set and the last sample set, it means that the first sample set is always predictive.
That's the entire argument. Players who "own" a pitcher or a pitcher who "owns" a hitter continue to do so. That's my entire argument and has been.
My ENTIRE argument is some hitters own certain pitchers and will do much better against them than against others. Similarly, some pitchers own some hitters.
It has nothing to do with 50 or a 100 or a 1000 players who are marginally above or below their average will go up or down due a myriad of factors. This has nothing to do with anything I have said.
My ENTIRE argument is some hitters own certain pitchers and will do much better against them than against others. Similarly, some pitchers own some hitters.
It has nothing to do with 50 or a 100 or a 1000 players who are marginally above or below their average will go up or down due a myriad of factors. This has nothing to do with anything I have said.
It's not a small sample. For many, it's their entire career.
Of course, most players don't own a pitcher nor do most pitchers own a hitter or hitters.
I ONLY talked about the ones that do. I NEVER talked about an average player. YOU CHANGED that not me.