GDT 4/10/18 Cincinnati Reds (Bailey) @ Philadelphia Phillies (Nola)

[Austinfan]

I dunno, are you talking #1 starter or HOF?

ERA+ 10 pitchers 149+, 15 pitchers 133+, 19 pitchers 123+, Nola was 23rd at 119
FIP, 12 pitchers 3.50-, 17 pitcher 3.72-, Nola was 9th at 3.27
WAR (BR) 10 pitchers 5.3+, 18 pitchers 4.2+, Nola was 16th with 4.3
WAR (FG) 8 pitchers 4.9+, 13 pitchers 4.1+, 20 pitchers 3.3+, Nola was 12th with 4.3

So what is a #1 starter? top 10 out of 30, top 20?
Same with a #2 starter? top 40, top 50?

[kikutaro]

I think the problem with discussing the concept of a #1 starter is that there are 2 common definitions. There is a scout based definition, which I will call "#1 starter stuff" and a fan based definition, which I will call "#1 starter results". For example, this link describes what "#1 starter stuff" means to a scout. To a fan "#1 starter result" means that his statistics are among the best. In Austinfan's post, he is talking about "#1 starter result", and Nola has very good results. On the other hand, andyb always talks about "#1 starter stuff", and per the above definition, Nola probably isn't quite there.

I think it would be good to point out whether you are referring to stuff or results whenever you use the phrase #1 (or #2 etc) starter.

[andyb]

So what is a #1 starter? top 10 out of 30, top 20?
Same with a #2 starter? top 40, top 50?

They are designations that combine stuff with performance and durability. That is why looking at statistical rankings alone do not show who is a number 1, 2, or 3. Saying Nola is a solid number 2 now is a compliment for him. There are very few true number ones. Not sure why you get bent out of shape by this distinction.

[julio]

There are very few true number ones. Not sure why you get bent out of shape by this distinction.

Probably because the assertion that, given 30 major league ballclubs, there are only 12 (or so) "Number One" pitchers seems...absurd? You have a definition that describes something quite rare, but you use terminology that implies something completely different (e.g., the top guys in 30 rotations).

For the unwashed: A "number one" pitcher is not the top pitcher on any major league staff. That's not what "number one" means; it means what andyb says it means, and there aren't 30 of them, or close to that many.

[andyb]

It is not my terminology. It is the terminology that scouts use. That is not absurd, it is just reality. There are many more #5s than #1s. The distribution is something like 10, 25, 40, 70, 100.

[Austinfan]

The terminology that scouts use is for pie in the sky projections, because they're talking about low probability upside, not realistic projections. Performance is all that matters in the Show, not potential or "stuff." The only value stuff has is projecting future performance, and even then, command matters as much as velocity (see Grienke).

While there may not be 30 "#1 starters", it's not as narrow as Kershaw, either.

Which is why I ask what the cutoff is, and durability is limited, how many pitchers put up 3+ seasons of "#1" performance. So you want enough performance to suggest a pitcher is reliable, but if 4-5 seasons are required, you only know who was a #1 starter in hindsight.

To me, the cutoffs are the median in each group, i.e., 15 #1s, 30 #2s (i.e. up to 45 pitchers), and so on.
Reason is you want all your pitchers to be in the top half of each category, so your staff as a whole is in the top third or better.

And since no one measure is definitive, you want someone who overall would be considered top 15 or so.
To me, Nola is there, he just has to do it for 30+ starts to confirm that status.

[andyb]

The terminology that scouts use is about stuff and secondarily about performance. Yours seems pretty arbitrary.

[julio]

It is the terminology that scouts use.

Couple of observations:

Jargon. It's not uncommon for jargon in any profession to consist of terms that don't mean (to the insiders) what they mean in everyday English to laypeople. That's one reason why people who want to communicate to lay people generally try to limit their use of jargon. It confuses people.

The terminology andyb uses undoubtedly is used by scouts - but not all scouts, unless scouting has changed considerably. Different organizations use different approaches, because they don't all share the same philosophy. Does a 5-point scale for evaluating pitching make more sense than a 10-point scale? Or a 3-point scale? Or something different. Opinions vary.

andyb is correct that, given his scale, there are a lot more #5s than #1s. That's true no matter what scale one uses, because talent in professional baseball isn't distributed normally - we're not seeing the full population distribution of talent, but only the upper tail.

If I wanted to evaluate pitching prospects, I think I would be inclined to use terminology - any terminology - other than a system that just happens to parallel the current professional baseball practice of having five-man rotations - unless I wanted to parallel actual usage for some reason. Did scouts use a 4-point scale before the advent of five-man rotations? A 3-point scale earlier than that? I don't think so.

I personally think that andyb's (and some scouts') five-point scale is probably too fine-grained to have real meaning. I note that we often can't talk about a pitcher as a "#5." He's a "#4-#5"; another guy is a "#2-#3." This is because the distinctions between these "grades" isn't clear. So why use a five-point scale, if it doesn't reliably provide meaningful distinctions? Most likely, it's used by many organizations' scouts simply because "that's the way we've always done it."

andyb is correct, I think, when he talks about something on the order of 10-15 pitchers who are in an elite class. He calls them "true #1s." I might just call them "Cy Young award candidates," and dispense with the insider jargon.

Aaron Nola? Don't know; when he puts up a season that puts him in the running for a Cy Young award (doesn't have to win it, but has to get meaningful votes), then the debate ends, as far as I'm concerned.

[Austinfan]

We're really not disagreeing, since if Nola put up 32 starts like last season and the beginning of this season he's be a 5+ WAR pitcher and in the Cy Young discussion. Which is why I say he has to do it for a complete season, but he's been pitching at that level since the first couple months of last season.

In the same way, Arrieta has to show he can pitch like he did the second half of last season with reduced velocity for a whole season to still be considered a #1 pitcher.

[zambonir]

This is the same as the "ace" debate, in some ways it is just semantics and in other ways it is not. All 30 teams have a "#1 pitcher" but probably only 10-12 of those guys would absolutely be the #1 pitcher on every team (and some teams have more than one of them).

In some ways I suspect we are also stuck in the past with this kind of thinking. Probably in the 1960s, when there were only 20 teams, and no free agency or massive payroll inequality, and four-man rotations with the odd guy who pitched over 250 innings, there were still probably just 10 or 15 "ace" pitchers (and some teams still had more than one). Now there are more teams, but maybe not more aces, and the older ones cluster around money.

I do think Nola could make a run at the Cy with a great season, though even in the age of analytics such things still correlate with counting stats, team success and reputation as well as performance. But are there currently 15 teams where, if Nola were traded tomororow, he would be that team's best pitcher, right now?