Hal14 wrote:sansterre wrote:Hal14 wrote:My point is that guy said AD deserves a vote because he's an insane 2nd option. But he's only been a 2nd option for 1.25 seasons which is not even close to a large enough sample size to base anything off of. This isn't a peaks project - this is top 100 greatest players of all time.
To be clear, I vote based on my set of formulas. AD is #2 in my formulas because the
blend of the different metrics I use (BackPicks BPM, PIPM, CORP, Win Shares, VORP, WOWYR and a manual playoff adjustment) likes him the second most. But if I post stuff like "I'm voting for AD #2 because my formulas like him the second-most. PEACE!" that may be seen as insufficient reasoning. So I basically try to put those things into words that I think will be persuasive to other readers. And the fact that (for a limited sample size) AD scaled really well next to another strong scorer will be persuasive to some other voters, so I include it, and trust their judgment to give it the appropriate amount of respect (whatever that might be). As with all limited sample size patterns, it means considerably more than it would if it had never happened, but but less than if that trend had continued for several seasons.
So longevity, durability, team success and intangibles like how much of a headcase they are, how coachable they are, good for team chemistry/team morale/locker room presence, being a good teammate - none of that factors in at all, huh?
It varies:
1) Longevity obviously matters. My method is to basically give a value to each season and add all the seasons up. More seasons = better. That's how Robert Parish is my #2 right now; it sure ain't because he had the 2nd best peak available. But if the seasons you *do* have are unusually good then you don't need as many of them for the same value. So longevity counts.
2) Durability matters indirectly. Injuries drop minutes (which hurts) and decrease performance (which hurts). Ideally I'd nuke every player's rating on a year where they missed the playoffs for injury but I don't actually have a very good system for that. So durability counts, but not in the sense of me giving a guy a bonus for not getting injured. Guys that play fewer games and miss more time rack up less value. But I'll admit that my system is almost certainly rewarding some players for seasons where they missed playoff time and I didn't catch it.
3) Team Success (in terms of on-court team performance with points scored/allowed) shows up in a number of those stats. But team success in terms of "how many rings" or "how far did they get in the playoffs" definitely doesn't matter to me unless there's a giant smoking gun (I'm thinking of pre-'67 Wilt here, a guy whose offensive stats screamed GOAT but teams adding him never saw their offenses budge much). Teammate quality is such a huge factor with this stuff. By 1988 Jordan's teams hadn't done crap, but it would be crazy to hold that against him. Andre Iguodala got a lot more attention with Golden State than with Philly, despite almost certainly being better in Philly, because in Golden State he was playing with way better teammates. There are times when team success *suggests* things that are worth looking at. If Bill Russell doesn't have the stat profile of a GOAT contender, but his team-level impact really pops, the stats may be missing something. If Draymond Green keeps leading strong defenses despite imperfect defensive teammates, perhaps he's better on defense than the stats can pick up. So, in general, I ignore team success because it's mostly teammates/coaching/era. But team success (in particular component stats) *can* suggest that there are things worth looking at. For example, I have T-Mac #1 despite his 'team success' being severely lacking. And I did poke around. His playoff stats got even better, not worse. And his teammates look comprehensively really, really weak (as far as BPM can tell). And both WOWYR and AuRPM think he was quite good (if not as good as BPM/VORP do), and both of those are driven by context and not by box-score stats. So I generally write off his lack of team success to really weak teammates and move on. Could be wrong.
4) intangibles like how much of a headcase they are, how coachable they are, good for team chemistry/team morale/locker room presence, being a good teammate . . . Not really. Obviously if they were a nightmare to work with you'd get weird team performance fluctuations, but it doesn't usually work that way. Plenty of players who are known for being really, really hard to work with and coach (Jordan, Kobe, etc.) are lauded for their winning spirit. I won't say that I ignore it. If Durant wrecked the chemistry of some monster GSW teams, he'll probably lose tie-breakers for me, but not a lot more than that. But most of that stuff is really, really hard to quantify. Everyone thought that Jordan was a cancer until he, you know, won anyways. That stuff matters, but a lot of it often gets back to team success. Being a bastard that wins means you've got swagger; being a bastard that loses means you're a selfish headcase. Having good teammates can skew a lot of that interpretation. I'm not saying that there's nothing there, but I don't really trust it well enough to rely on it the way I rely on other things.
