Custom Valuator Problems

[rotonut]

I said this in the other thread and I'll say it here. The CV has been quirky and unreliable since its inception about three years ago. I personally have constructed a spreadsheet to calculate my own values. That the CV spits out values that make you scratch your head is enough for me not to use it. I need to bring something to the draft table that I know is realiable. I don't need to be thinking about what the programmers had in mind or why the values seem off. I need good, workable values.

With that said, my suggestion is to use the RotoLab engine as HQ's cutom valuator. They are already in a partnership with respect to using HQ's projections. Why not take the next step? I don't see any threads questioning the RotoLab values as I do the CV values. I know it's too late for this season, but please think about it for next season.

I am also a long-time subscriber and have remained a loyal subscriber because the content is great, but this "tool" is just too unreliable. Plus, way too much time has been devoted to fixing something that should not have been rolled out in the first place if it had quirks. We should be talking about auction strategies, not why the CV doesn't make sense.

Andrew

[BEF]

Something is still not right with this program. Using the 3/25 projections file, I customized it for a 14 team mixed 5x5. Take a look:

PIAZZA (384 AB's)
.263/18/54/0 value of -$6.63

BONDS (159 AB's)
.333/18/41/2 value of -$12.69


I know that the AB's difference is huge, but does that really matter THAT much? They have the same number of homers. Bonds is a little short in RBI's but blows away BAVG, and has 2 extra SB's. How could he be priced like this? How can he go from a positive $13 in the NL only spreadsheet to a negative $13 in a mixed league?

What's going on?

[RAY@HQ]

I'm not convinced the negative valuations are accurate... as I said earlier in this thread or another, the key is to note that the guys ARE negatively valued, and therefore are not optimal plays in your league.

The problem with Bonds now is with the ABs. A .333 BA over 159 AB has just about no impact on your overall team BA, which will get amassed over something like ~5500-6000 AB. His other numbers are no longer valuable in that little playing time either, relative to what else is available in your shallow league pool. Bonds is, in short, producing at a lower rate than a lot of other options, thus the negative valuation.

As I said, it's not overly important to see HOW negative he is, just that he IS negative.

[maisany]

USAToday CV

Has anybody seen this?

The layout is eerily familiar.

[BEF]

Here's my concern: I'll be using RotoLab during the auction and will be looking at every owner's Profit and Value numbers after each player is purchased. And that is what is severely affected by this issue; there's a big difference if I acquire a player who's valued at -$1 versus -$13. I know, I know, negative is negative, but if you're paying attention to Profit (in order to get to that magical $320 in Value), then this issue really comes into play.

I don't believe the answer is simply "don't get anyone with negative numbers", because there are in fact a few players that I'm willing to gamble on due to positive notice at HQ. For instance, Keith Ginter is listed at $8 in the standard single league projections, but he's at -$10 in my 14 team mixed. So if/when I get him, my Profit/Value takes a real hit, and skews all my team data, even though he's been touted at HQ.

Not sure if this aspect of things has come up elsewhere in the various threads about the CV; if so, I apologize and would appreciate being directed to that thread.

[RAY@HQ]

BEF, am I reading right that you're somehow integrated Custom Valuator numbers with Rotolab? Isn't that redundant, since Rotolab does the custom valuation for you, separate from our CV tool?

[RotoMaster]

Re: USAToday CV

Originally posted by maisany
Has anybody seen this?

The layout is eerily familiar.

They use the engine and projections from RotoTimes. If you compare the values they are identical.

[Steve NY]

Another Valuator Question

Excuse me if this has been addressed previously.

I understand HQ's justification for negative values. What confounds me is that the negativity for hitters in my league (11 tm NL, 5x5, $260) "starts" at approx. $-7. That is, nobody receives a value of between $-1 and $-6.

The negative valuations for pitchers have a more reasonable flow...there are pitchers valued at $-1, $-2, $-3, etc.

How is it possible that no offensive player is judged to have a value between $-1 and $-7?

[RAY@HQ]

That's been noted a few times, Steve, and I don't have a good answer for you. My advice has been to note which players are negative, but don't sweat over "how" negative they are.

And for anyone reading this thread, if you haven't seen Ron's note on the home page (posted Monday), check it out.

[BrooklynMonster]

95% accurate?

I haven't nor do I intend to do and extensive survey of the custom draft guides results. I pay you guys to keep it accurate and reliable, but I'm still seeing anomalies that discredit the entire tool.

Hideki Matsui; 91 runs 28-97-.296 2 sb's

Sheffield; 89 runs 28-95-.295 4 sb's

These projections are nearly identical yet Matsui is valued at $27.43 and Sheff at $22.11??? More than $5 less? This is 95% accurate? I seriously doubt it.

[RAY@HQ]

The issue in this case is batting average... Matsui's comes with ~60 extra AB, which is giving it extra weight. Now that's theoretically accurate, Matsui's .296 for an extra 60 AB should be worth a little more than Sheff's .295 for 60 fewer AB. But I agree that a $5 difference is dubious, and we're still working on that.

Just to prove that this is the issue, re-run your numbers without BA as a category... Sheff and Matsui's non-BA contributions are valued within a few cents of each other, which is as it should be.

Calculating the value of BA is a 2-variable calculation: what's the average, and for how many AB do you get it? We're still working to get that valuation nailed.

[BrooklynMonster]

thanks

for the prompt and clear explanation.