What "values" do you want for use during your auction?

[alleyoops]

With the discussions about the CVRC and Rotolab, it got me thinking what numbers would be most useful to me during an auction? Sorry, I don't to any draft leagues, but perhaps the discussion is similar for those.

Ones that come to mind would be:

1. Projected values - mathematical "value" on each player based on a set of projections and some value methodology.
2. Inflated values - same as 1, but taking into account keepers and players already bought in the auction so far.
3. Expected league bid values - starts with 2, but incorporates league tendencies.

Are there others that would be useful? Of these (and others that are suggested), which would be most useful to you during an auction?

I order them 2, 3, 1, but think each would have some value. ln actuality, I've used kind of a hybrid of 2 and 3, which Rotolab enables me to do, in a crude way. It sounds like CVRC would allow a combination of 1 and 3, or maybe 1, the keeper part of 2, and 3 with some work.

The problem with using just 1 is that you haven't adjusted for inflation, and you don't know what the guy will likely cost you.

The problem with using just 2 is that you don't know what the guy will likely cost you.

The problem with using just 3 is that you risk rostering some players whose salaries are higher than their "real" value. (i.e. it "pushes" you toward the league tendencies)

[viper]

This season I will be four straight drafts and 4 auction drafts. I think I actually want #1 but with certain pitching category values having less or more weight based on my league's past tendencies. I go in with paper so making on the fly adjustments are done based on my sense of what is happening. I want to know a player's general value based on what they bring to the table and then make my own adjustments as the draft progresses.

Part of my going in position is there will be more $1 players than any list provides. This gives me some dollar leeway for the top flight players or, at least, that is the way I want to look at it.

[cwk1963]

I want #1. 2 & 3 I can figure myself as I see the auction unfold.

[rotonut]

As I'm sure most of you are sick of hearing, I contend that #1 already incorporates some of #3. That is, the value methodology already incorporates some league tendencies, i.e., hit/pitch split and category weighting. In a perfect world, I'd want all three values so that I could adjust mid-auction based on league tendencies.

[Guest]

The truth here (I think) is what we'd all really, really want is a methodology which "learned" your league during the auction - that is looks at the prices paid for each player and uses that information to calculate what players "should" go for based on that information as the auction goes on based on $ remaining and previously valued players.

I have messed with this a little bit with varied results. The reality is you need a somewhat hybrid SGP/REP system to do it (yes Viper, thinking of you here), and the risk here is that you couldn't blindly use the values - you'd have to think it through, which some want to do, and others don't want to do.

However, you guys should keep working through this - what are you trying to measure here? Values which mimic your league would have value in that you wouldn't overpay players in the context of your league supply/demand but might not lead you to the optimal result (if everyone in your league jumped off a bridge....) I would think your ultimate goal would be values which help you achieve your strategic goals, whatever those were.

[alleyoops]

The combination of 2 and 3 that I used last year is kind of like what you're saying, Gary. It starts with a full set of projections (from BBHQ, but you can edit them). You can specify a valuation method (PVM, SGP) and vary the associated weights for categories. You can also set two "sliders" - one for "stars and scrubs" - essentially flattening the value pool to a varying extent at the top and bottom - the other "spread the risk' - doing the opposite - steepening the top and bottom to move values more toward the middle. By varying all of these, you can try to reflect the league bidding tendencies, if you so choose. Of course, if the league has a bias toward young players, or away from injury-prone players, for example, you'd have to edit the projections on these guys to reflect that bias (unless the projection guru already factored in the same biases).

The result is a league-tendency-oriented set of values. It contains the projection site's biases, of course, but also reflects the league bidding tendencies. Then, as a player is bought (or kept), the software recalculates these values "on the fly". If the player went for less than the previous value, that difference is now reflected across the board in other player values. In truth, one player being $10 more, or less, than value at the beginning of the auction isn't going to impact the other values much at all. But if you load in (say) 30 keepers, all of whom are $10 to $15 below the prior values, you'll see an impact (inflation, essentially).

While it isn't "learning" league tendencies from this particular auction, it is using historic tendencies, to whatever extent you want it to do that.

One of the odd nuances of this is that in the latter part of the auction, you see some real strange values. If a guy who was initially valued at $15 has somehow not been bought yet, and a few owners have a bunch of money left to spend, you might see a calculated value for him of $50! This reinforces your point that you shouldn't necessarily bid to these calculated values. On the other hand, if you're one of the guys with a lot of money left, you probably should.

Another feature of the software is that you can display this "inflated value", or you can display the original "uninflated" value. That's part of the reason for my question in the first place. I'm not sure the inflated values are as useful as the uninflated ones, and I'm not sure I should be tweaking those weights, etc., to match my league.

[Guest]

I think I'm describing something a bit different, but I do understand precisely what you're talking about.

I actually am not a huge fan of inflation as it has been applied to the fantasy industry, but I of course believe in it using a different application. I am completing a rather large site piece and then plan to describe my thoughts on this in a following piece, but basically I of course understand what happens when there's $300 of "value" remaining and $100 left to spend.

For example, and I've shared this with Todd, I think the 65/35 split is the result of two factors:

1. You don't need to buy as many stats to win pitching categories as you do to win hitting categories (as a percentage of the player pool)
2. There is rampant disagreement in the pitchers who compromise the bottom 10-20% of the pitching pools of leagues of most/all types (that is, everyone agrees on the top/mid-range SP and the closers) and then after that, the bottom of the pitching pool mostly goes for a buck or three because half the lists don't have the player as positive value. I think this is much different in deep leagues for hitters (where MOST of the hitters are in everyone's pool because there probably aren't 168-182 fulltime hitters in those leagues).

In shallower leagues the bottom of the hitting pool is more foggy but so is the pitching pool, but I'd actually think this supports to some degrees why the split is more pronounced in only leagues than MIX (or it's been proposed that way). I don't think it actually has much to do with the top pitchers in the league, and I tend to think they go for a discount in leagues using any traditional valuation system.

Back to the point of the question - Im much more interested in real time calculations of what a league is paying than any type of historicals, but if you are really comfortable with it, obviously with some excel time you can reproduce what your league can do for the most part. Then you just have to figure out if you're going to follow along.

[roche]

By varying all of these, you can try to reflect the league bidding tendencies,

In my experience, this is barking up the wrong tree. What I'm saying is that there are no league bidding tendencies.
In an auction, the price for any player is set by the last 2 owners bidding on that player. Lots of people may participate in the bidding initially but near the end, it always comes down to 2 owners facing off for the last few bids. It's rare that there are multiple bidders all trying to win the same player right to the gavel. For example, if you have 2 owners in your league who both love to take chances on rookies, these 2 will bid each other up for each and every rookie. The rest of the owners drop out of the bidding fairly early and sit out the rest of these silly bidding wars. This then is the supposed bidding tendencies for this league. In truth, it's just the bad strategy of 2 owners.
As well, consider the tactics of winning. By definition, the winners of the pool are the teams that buy the most stats with the $260. In other words, the winners make the most profit out of the players that they buy. How do you make the most profit? Buy under-valued players. Avoid bidding wars. Don't buy your favorite players. Anticipate the market. Zig when others are zagging. Winners don't have bidding tendencies.
Now, consider the teams that don't win. If last year, I bought 2 prime-time catchers and my team sucked, how likely am I to want to buy 2 prime-time catchers again this year. If I went with no closer last year and lost, I'm probably going to buy at least one this year. No one wants to repeat a losing season. Everyone thinks that they can learn from their mistakes. (Yes, there are owners who make the same mistakes year after year and lose year after year but you don't really need a fancy computer program to beat these people.) This all adds up to each team changing their bidding tendencies from year to year. Therefore, there really is no bidding tendencies.
Actually, there is one bidding tendency that is pretty universal and that's the local team inflation. Players from the local team always cost more than they should. However, in mature sophisticated pools, this is usually just $1-3 more so the effect is minimal.

[rotodog]

During a draft I dont need anything other than Actual value. I have always had at least one set of Non adjusted Valuations. Whether on paper or computer. Then I usually had adjusted valuations to my strategy somewhere handy. If i adjust, it was only on the top level talen that I am concerned with..

For instance, one league I have been in has a 280 Budget . NL, Auction keeper with massive inflation. 12 teams.

In the past, we have had 11 owners and to keep the integrity of the dollars and players values, we have added an extra UT and an extra pitcher and take the BUDGET up to 305 . In this situation, it changes values, but the extra money tends to get spread out. So, there is 25 extra dollars per team. Do you buy 2- $12 players? Or do you go an extra buck on 23 players? Thats what others do......But not me

The extra 2 player slots have an extra 25 bucks to be spent on them. For me the extra pitcher will be a buck and the last hitter will be a buck too, so I essentially have 23 extra dollars on draft day to be spread through my team.

I would rather go 5 bucks more on 4 studs, or 10 bucks more on 2 MEGA STUDS..... Others will make the mistake of running that extra 25 bucks per team through rotolab and spreading it out. I would make a mental adjustment, but if I was using rotolab, I would tweak the Stars and scrub slider HARD to the right until I saw the 23 extra dollars show up in the difference between the NON adjusted and the tweaked...

For instance: Hanley Ramirez 46 bucks
Tweaked: 52:
Tweaked inflated: 62

he will go for 56 bucks. If I drop a 53 bid to start, usually people shuffle like crazy because they cant believe what i just did. Some stuttering and shuffling later, someone says 54. I jump bid to 56 and its over....

At the draft table , I took the 23 bucks and gave 10 to Hanley over Standard, NON tweaked, non inflated value.

Ive been successful with strategies like this and why i prefer AL/NL only leagues or leagues with non standard categories. It Allows for manipulation of values for an advantage.... But I do understand its not a perfect science....

[roche]

I would rather go 5 bucks more on 4 studs, or 10 bucks more on 2 MEGA STUDS..... Others will make the mistake of running that extra 25 bucks per team through rotolab and spreading it out. I would make a mental adjustment, but if I was using rotolab, I would tweak the Stars and scrub slider HARD to the right until I saw the 23 extra dollars show up in the difference between the NON adjusted and the tweaked...

For instance: Hanley Ramirez 46 bucks
Tweaked: 52:
Tweaked inflated: 62

he will go for 56 bucks. If I drop a 53 bid to start, usually people shuffle like crazy because they cant believe what i just did. Some stuttering and shuffling later, someone says 54. I jump bid to 56 and its over....

At the draft table , I took the 23 bucks and gave 10 to Hanley over Standard, NON tweaked, non inflated value.

Ive been successful with strategies like this and why i prefer AL/NL only leagues or leagues with non standard categories. It Allows for manipulation of values for an advantage

What you didn't say but is pretty crucial to your success is that you are still able to buy your "usual roster" for the same un-tweaked prices. So you end up with your usual roster and a bonus of Hanley. This does not work in a super league with proper pricing and tight price-enforcing so that you miss out on all of your usual players. Spending extra on stars is great as long as the scrubs that you end up with are of acceptable quality. In the end, the goal is still to buy the most stats for your $260 (or $305). Overspending on studs just means that you better find lots of bargains later. Mind you, there is no such thing as a "super league with proper pricing and tight price-enforcing" so this strategy will work pretty much all of the time.

[rotodog]

I would rather go 5 bucks more on 4 studs, or 10 bucks more on 2 MEGA STUDS..... Others will make the mistake of running that extra 25 bucks per team through rotolab and spreading it out. I would make a mental adjustment, but if I was using rotolab, I would tweak the Stars and scrub slider HARD to the right until I saw the 23 extra dollars show up in the difference between the NON adjusted and the tweaked...

For instance: Hanley Ramirez 46 bucks
Tweaked: 52:
Tweaked inflated: 62

he will go for 56 bucks. If I drop a 53 bid to start, usually people shuffle like crazy because they cant believe what i just did. Some stuttering and shuffling later, someone says 54. I jump bid to 56 and its over....

At the draft table , I took the 23 bucks and gave 10 to Hanley over Standard, NON tweaked, non inflated value.

Ive been successful with strategies like this and why i prefer AL/NL only leagues or leagues with non standard categories. It Allows for manipulation of values for an advantage

What you didn't say but is pretty crucial to your success is that you are still able to buy your "usual roster" for the same un-tweaked prices. So you end up with your usual roster and a bonus of Hanley. This does not work in a super league with proper pricing and tight price-enforcing so that you miss out on all of your usual players. Spending extra on stars is great as long as the scrubs that you end up with are of acceptable quality. In the end, the goal is still to buy the most stats for your $260 (or $305). Overspending on studs just means that you better find lots of bargains later. Mind you, there is no such thing as a "super league with proper pricing and tight price-enforcing" so this strategy will work pretty much all of the time.

That it does my friend. I just hope anyone in my league isnt here reading this....

But that is the premise. In a keeper league it is much easier to employ this strategy with keepers at under value many times too....Thats where the inflation comes in... You certainly do need the abilty to find the right scrubs, but it forces me to every year and every year I dont disappoint... I Always find a Skip Schumaker type for a buck that gets me Abs....Always...

One year I might get bit in the A$$, but it hasnt happened yet...

But the league example above is just one way to show that values themsleves do not win leagues. Its values used in the right context and some luck that wins!

[davep]

When using the CVRC, what values is that giving you? Are those suggested bid values, or expected auction values?

[rotodog]

When using the CVRC, what values is that giving you? Are those suggested bid values, or expected auction values?

They are projected actual value based on projections calculated by totaling the players contribution to each stat pool....

So it is neither really....Suggested bid values and projected valuations are not the same thing...