Mon Jan 09, 2017 10:44 am
I agree that the whole set needs to be repriced. The current prices are a mish-mash of several different pricing models, and all of them suffer from the crazy idea that players should be priced for a "neutral" ballpark. This is a big factor in your "power creep".
Ballpark homers are the single biggest variable for hitter cards as the ballpark changes. For pitcher cards, surprisingly enough, it's ballpark singles - because the community quite rightly shuns guys who give up a lot of homers.
But there are four distinct games going on here regularly - uncapped (or >$200 million capped) live (and occasionally auto) drafts, $140 million and $200 million leagues, the nuts and bolts caps ($60, 80, and 100 million) mostly autodraft leagues, and the franchise leagues which usually have funky caps in the $100 million range. Pricing for these leagues has different needs.
For the very large caps, pricing of the cards should be more or less in the order that the cards come off the board in live drafts. This creates a problem; Bill Dickey, Joe Mauer, and Arky Vaughan, as LH bats in positions where RH bats are overwhelming in number, get drafted out of all proportion to their card values by the guys planning to play in LH parks. A similar problem afflicts LH starters in reverse, they get pushed way down the draft because of Juan Gonzalez, Cecil Fielder, Josh Gibson, Jeff Bagwell, and the like.
This wouldn't be a problem, but then there comes the issue of how and when do we merge the values of lesser valued LH bats and LHP into the mainstream pricing model. If we have a discontinuity, everyone will howl. If we try to do it less noticeably, we create a small group of LH starters who are serious bargains starting at about $140 million and running down at least to $100 million and the franchise leagues. And LH hitters would have the reverse problem, some of them would be seriously overpriced at moderate caps.
Another issue, that mainly affects the low cap leagues, is what should be the shape of the NERP vs. price curve. The problem can be illustrated as follows:
The best Ruth card is worth roughly 38 NERP in a particular ballpark, Bryce Harper in that same ballpark is worth about 23 NERP. The difference in their salaries is almost $5 million. Fine, we will price the cards at 3 NERP = $1 million. Harold Baines in that ballpark is worth about 8.4 NERP and has a salary of 6.21 million, not too bad. Everything is good so far,right? Now the rot sets in. According to this model, a 0.50 guy should have a NERP of -9 or so. But there are about 400 cards in the set with NERPs less than -9, and only about 175 or so 50 cent guys. Somewhere we have to break the linear model.
Where should we do this, though? One of my quick guides for setting up a team is to assume that I will have 8 guys whose main role is to not chew up salary. This consists of my throwaway 50 cent starter, 6 bench guys, and my mopup reliever. For my other 4 relievers, I usually allocate 1.5 player salaries. Now I have a starting 9 and 4 starters to pay for along with my 1.5 salaries to the pen. I knock about 5 million off the cap, divide by 14.5, and that's my average salary for my lineup and rotation.
For the 60 million cap, then, I'm looking at 4 million per salary slot. for 80 million, it's about 5.4 million, and for 100 million it's close to 7 million. So I think that cards from about 2 million to 9 million should be priced to cater to these caps - this will in general require a different linear slope than we generated from Ruth-Harper. But some of those 9 million players get used in the uncapped leagues, and they need to be priced (In my opinion) by draft order rather than NERP.
We can handle this by making the price vs. NERP curve slightly concave, 1 NERP in the top range is worth slightly more salary than in the moderately priced cards. This happens in the current pricing model, intentionally from what I gather.
But we are still left with the sub-2 million but over about .75 million cards. If we price those on any model they are only going to be used if they are seriously lopsided or have some redeeming factor such as a terrific base stealing number or primo defense at several positions. And the scrubs, from what I can see, should be priced at about a penny per NERP which will totally change the way the bottom cards are valued.
Fine. Are we done yet?
Not quite. Now we have to deal with platooners. How should Gates Brown be priced? How about Kal Daniels? Or Sean Berry? Pricing any of these guys by giving them, say, 2/3 of the value of their NERP vs. RHP and 1/3 of their value vs LHP, leads to making platooning a hugely successful strategy. (Fortunately, HAL doesn't handle platoons very well, or nobody would ever use a balanced hitter.) But what's a fair way to price these guys? I have some ideas, but they are hugely complicated and not really very successful. It's also relevant to note that platoons don't work as well in lopsided parks or in the total bomber parks as they do in Petco or the Astrodome.
Whew! Pitchers have got to be a snap, right?
Wrong!. They're worse. Pitcher values are hugely dependent on their home ballpark, and are also significantly influenced by the other ballparks in the league and the hitter mix they will face. And then there's the issue of no DH leagues - how do we price the pitchers who can hit? If we ignore this then almost any 4, 5, 7 or 8 rated hitting starter is a huge value. If we price those guys for (most of) their value as a hitter then they're unusable in DH leagues.
Then there's the value of ballpark homers on pitcher cards. In lower caps, people use more "W" hitters than in high caps, so obviously the ballpark homers are less of a penalty for the lower valued pitchers. I proved this, more or less, by using a starting rotation where my highest priced starter was Jeff Brantley and winning a ring at $80 million. Of course, I had a hell of a bullpen and "quick hook" going for me.
Pitchers also have BK and WP ratings and Hold ratings. How do we model those? Again, these things are vitally important at high caps and somewhat less important at $60 million. Do you have a good model for that? I don't.
Another slippery topic for pitchers is what to do about extremely lopsided pitchers like Cone, Caldwell, Ontiveros, and Franco. I think you have to handle lopsided relievers totally differently than lopsided starters, since for the starters, your opponent gets to decide who hits against them while for relievers you get to (try to tell HAL to) decide who they face. I don't think the current card set is priced that way very effectively.
And then there's the issue of RP pricing and its cousin, the pricing of S/R guys. Obviously, we have to use a different scale here. If Eric Gagne were an S9* he'd be worth what, 13, 14 million? I haven't looked exactly, but that seems right. He'd be throwing 330, 340 innings a season. As an R1 closer, it's hard to get him to 80 if you try, although I saw a guy get him to 130 or so a while back. And don't get me started on Dale Murray. Don't you hijack this thread for him, either, there are other guys with the same issue. So how do we price relievers? It should be a different model for high caps, where even Murray and Gagne are going to get lit up a couple of times a week, vs. low caps where they might get lit up 3 or 4 times a season.
If you have any ideas about how to make reliever pricing fair over the whole spectrum of leagues that are played here, be sure and let me know. I have some kluges that sort of work, but I'm not happy with them and tinker constantly.
Finally, let me explain why pricing the cards to a neutral park is silly.
1. Nobody uses neutral parks much, since it's very hard to win in them. The reason for that is that those parks don't favor or disfavor any particular type of card, everybody does fairly well in them.
2. Pricing power cards to a neutral park makes power hitters extremely cheap in Wrigley 78 or Fulton or the like.
3. Pricing singles hitter cards like Cobb and Speaker to a neutral park makes them more or less unusable in power parks AND DOESN'T REALLY HELP THEIR value in Forbes or Royals. Pricing power hitters and small ballers to the same ballpark makes small ball very difficult.
4. Pricing lopsided pitchers to balanced parks greatly understates their value in lopsided parks that favor their cards. This isn't, in general, true for lopsided hitters because of platooning. Similarly, pricing pitchers with HR troubles to neutral parks greatly understates their value.
I think the best way to price the card set is to generate 7 sets of card NERP values for these ballparks. (all zeroes, all 20's, all 10's, 0/0 & 20/20 (Polo), 20/20 and 0/0 (Forbes more or less), 0/20 & 0/20, and 20/0 & 20/0.) Then average the top 3 NERPs for every card and use that for the final card NERP. But that's a hell of a lot of work. I could probably do in a week or so using the tools that I have made. If a couple of other guys could generate a similar set of NERP values, I'd also be happy to merge the data sets together and see what we come up with. But the NERP values would have to include, as a minimum, defense for hitters and some adjustment for pitcher hold, BK, and WP. I could deal with the hitter issue separately, that's not an issue that any of us is going to have weird numbers for.
