It’s a mathematical equation that gives results. That doesn’t infer anyone’s opinion about the results and I would doubt there are many people if any who would agree Palmer is the inferior pitcher.

James did, in actual fact, select Perry over Ryan. Whether one agrees with that choice or not is an opinion, which can be debated. The fact that it happened is not an opinion, and is not debatable. Pretending James did not select Perry over Ryan is just blatantly lying.

One can admit the actual demonstrable fact, that James did in actual fact select Perry over Ryan, without arguing that this choice is one they agree with. One can reasonably argue for Ryan, it does not require lying about the record. Over the top claims quickly proven to be factually false never help an argument.

I agree most people would think Palmer is better. But that is different from saying the developers of the equations don't believe in them. These are very different propositions.

Not to justify Reuschel over Palmer, because I would probably take Palmer myself although I haven't studied it, I've certainly read people's views that Palmer benefited disproportionally from having some of the best defenses in history playing behind him. Have not seen any detailed analysis though.

If James developed his rankings on statistical parameters and not what he thinks about each player then no, he didn’t choose anyone over another anymore than the BR parameters did. BR presents a list of its data-run conclusions and does not offer any inference about what an individual person thinks.

James' list is what he thinks about each player, with hundreds of pages written alongside his listing in numerical order of he selects over who. A list of players sorted by WAR is not the same thing and has absolutely nothing to do with James list, which his total evaluation. Stop lying.

Obviously you have no familiarity with James' book.

The very reason people develop statistical analyses is because they believe that traditional subjective rankings are inadequate. So to say they don't believe the results of their analyses is senseless. Might one find isolated extreme examples where the developers would agree that their model didn't yield the right result? It's possible, any model has limitations. But your whole, oh that's just a ranking, it doesn't have anything to do with who the developers think is better schtick just makes no logical sense at all.

Right so theres Jim Palmer and Rick Reuschel and according to BR the choice isn’t even close as Reuschel is 11 spots higher. But I wonder how many people would prefer him on their team.

James even called out in his ranking comment for Ryan that he gets this high only because James believed a starting basis of zero was the appropriate thing to use - and that if measured against league average instead that Ryan would suffer significantly in the rankings. The implication was that he was slotting Ryan about as high as he could justify possible, a bias to him.


It's not hard to acknowledge that Perry and Rayn are pretty similar for career value but to argue that you would take Ryan. It does not require a long series of absolute lies and pretending using math (which is invalid unless you originally wrote the equation, as we have learned) is a problem lol. The Ryan fans could easily and reasonably pick their guy, but have steadfastly refused to even approach the realm of the reasonable or discernible reality.

I don't understand the BR models well enough to know why Reuschel comes out higher than the traditional assessment. I suspect it's a combinaton of things. I always wonder the same thing about Grich.

It’s just a list using JAWS-S, JAWS adjusted to punish 19th century pitchers. Balancing career WAR and WAR 7.

Do you have a sense on why WAR is kind to Reuschel?

One of pitcher WAR’s big components is the runs expected for a replacement pitcher - relying on that teams defense. WAR has his teams being worth negative defense, making a big gap between him pitching very well anyway and the projections based on the data that he had a horrible defense behind him prone to giving up many runs on balls in play. WAR also factors how runs were scored - rewarding FIP components. Reuschel was good at avoiding the long ball. He was good at keeping batted balls in play vs. out of the park, but even with a really bad defense and that tendency, he still did a good job at not giving up many runs (pretty similar ERA’s to Perry and Ryan), an impressive record.

The other component is that Palmer gets punished for these same things. Palmers ERA to FIP is an unusually huge gap, because FIP measures these same things. He has a 2.86 ERA, but a 3.50 FIP which typically demonstrates over a long sample lots of luck or a really, really good defense, which I don’t think anyone would dispute Palmer had behind him.

I’m sure there’s more smaller factors too, but this is the root cause.

Belanger and Brooks are possibly the best left side of an infield ever. Paul Blair was considered a great center fielder.

I don’t think I can think of a better defensive team overall off the top of my head than those Orioles clubs.

You can see Palmer’s results were not the result of luck by checking the team totals back then. The Orioles with Palmer off the mound produced highly unusual gaps between FIP and ERA because Weaver had such a great defensive squad out there.

My personal objections to WAR are that one has to agree with tons of valuation values to find it correct, but more significantly the root concepts that 1) the true actual result doesn’t matter much and 2) the bar of comparison should be a fictional made up minor leaguer instead. I liked that Win Shares was based on real world actual wins, and metrics like ERA+ that do great at contextualizing without losing that connection to real world event. All the same, this is a good example of why these metrics are valuable - I don’t think I’d say Palmer was worse but WAR has really opened the door to showing that Palmer had a massive amount of advantages people weren’t really factoring before and that Reuschel has been unfairly overlooked and did very well in a very difficult context.

Peter,

Sorry I must have missed that, when working my way through the pages of discussion.

Statistically speaking I believe the advantage is Perry. Could I fault someone for taking Nolan Ryan though? No. Both were extremely dominant pitchers, Ryan from a pure power standpoint is tough to beat, but, I think you give it to Perry though for the consistency that he managed over the course of his career though. I think Ryan would fair better in todays game with how velocity has trended and his durability, but this is a moot point since we can't plop players from one era to another.

Agree. I could see it going either way, but I certainly don't think it's self-evident that Ryan was better and it's certainly not true that no one would think otherwise. People do think otherwise!!

Agreed! I think another pitcher we could easily throw into this discussion, both in terms of comparability and being undervalued, is Phil Neikro.

I saw Niekro in Boston towards the end of his endless career. He had nothing that day. Baylor hit one off him that must have left the park in 2 seconds and gone 1000 feet.