The Dynamic Valuation System

Nine Bo Jacksons reader Jordan Lyall recently sent me an email containing a spreadsheet he designed that merged live scoring with IPDS' categorical dollar values.  His spreadsheet tracked the dollar values for players as they were selected by each team, and also provided a comparative analysis of all the teams in the draft.  The live scoring was helpful because it showed what holes my team was developing as I was drafting, and made me give certain players a longer look.  As an example, around round 14 in a mock I was doing, I noticed that I was significantly behind the other teams in stolen base values [a common characteristic of my teams].  That certainly made me look longer at Willy Taveras, who helped to bridge the gap in that category.

Jordan's spreadsheet pushed me to create what I'm referring to as the Dynamic Valuation System [DVS].  The logic is very similar to what Jordan created, but I made a few modifications.  Instead of tracking teams by their total categorical dollar values, I thought that DVS should track teams by their actual statistics - providing real time statistical standings.  Furthermore, I wanted to quantify a concept I knew existed for a while.  In the previous example, I knew Willy Taveras would help me in SBs, Rs, and AVG, but I wanted to know if those gains would outweigh the loss I would take in HRs and RBIs. I used a concept that many who are familiar with Art McGee know about: standings gain points.  To quantify this, I calculated what place in each category my team would be with Willy Taveras, and subtracted the points I already had in those categories.  For example (using made up numbers), I calculated that he would move me from 5th place in steals to 10th place, meaning it was a +5 gain in that category.  I did this for each of the categories, and used the sum of those values to represent his SGP value: how many points would Willy Taveras move me up in the standings?  I would then look at other players and decide which player would get me more points in the standings. 

In the few mocks I've done with DVS thus far, it's helped me prevent a problem of mine: statistical overdrafting.  I tend to be dominant in the rate stats, and lack in power and other counting stats.  DVS has helped me correct that by telling me that adding another .300 hitter in the 17th round doesn't help my team if I'm already in 1st place in that category.  You don't receive a 13th point for having an average .010 higher than 2nd place, so don't draft that .300 hitter if you can gain more points by drafting Khalil Greene, whose power numbers will help more than his AVG will hurt (in that situation).

When you open DVS, you'll notice a few things.  For one, it is much smaller than IPDS was - this one weighs in at a lean 12.5 MB.  Also, it is 100% macro free, which means that a lot of the problems some of y'all had with IPDS will be nonexistent.  Instead, it uses autofiltering, much like Tim Dierkes' RotoAuthority projection sheets.

So, how do you use this thing?

• When you open DVS, you'll go to the 'TeamNames' tab.  Here, you'll input the names of all the teams in your league.  This is important in tracking the picks as the draft goes along.
• Once you've inputted all the team names, go over to the 'DraftSheet' tab.  By default, you'll only see six columns.  You can unhide the columns to reveal each players statistics and categorical SGP data.  I set it to six columns in order to keep all of the vital information on my small screen at the same time.  You can expand and contract it as you'd like.
• As each player is selected, click on their corresponding cell in the "Drafted By:" column.  Click the arrow that will show up, and click the name of the team who selected the player.  It is absolutely vital that you do this correctly.  If you don't, the standings and SGP data will be thrown off.
• On the 'Standings' tab, you get...you guessed it, the current standings of your draft.  Also, you get the actual statistics below the standings board.  The 'Hcount' and 'Pcount' will track how many hitters and pitchers have been selected by each time.
• When drafting players, I wholeheartedly encourage you to learn how to use the AutoFilters.  Although I don't have time to give a tutorial, I can tell you that learning how to use them makes a big difference.  As an example, if you're down to the last 3 rounds and only need to see catchers and relief pitchers, autofiltering can do that for you.

Now, time for a few warnings.

• By default, you'll notice that every hitter has a SGP value of 55.  Why does A.J. Pierzynski have the same SGP value as A-Rod before any player has been selected?  Because even A.J.'s .266 AVG, 14 HR, 55 R, 58 RBI and 1 SB will move your team to first place in all of those categories, as they are all better than the 0 that the rest of the league has.
• You'll also notice that the starting pitchers have a SGP value of 22 when you start up.  Why?  Because even Johan Santana's 3.07 ERA and 1.06 WHIP is going to be "worse" than the 0.00's that the rest of the league has.  This leads me to...
• Don't get carried away with SGP data early on.  Given the actual nature of drafts, you won't get good use out of SGP data early on, as evidenced by the A.J. Pierzynski example.  Draft with dollar values, ADP, or gut feelings early on.  The SGP data really makes the difference in rounding out your team after round 10; not creating it from round 1.
• If players have the same SGP value, use the dollar values as the tiebreaker.  You may find yourself in an interesting situation where a -$10.00 pitcher will have the same SGP value as a $4.00 pitcher.  It's unlikely, but possible.  Do the smart thing and take the player with the higher dollar value.
• I recommend keeping the sheet sorted by ascending ADP.  This helps you avoid drafting someone 5 rounds before you should.  As a general rule, I'll only reach one round for a player in rounds 1-3, two rounds for a player in rounds 4-10, and three rounds from 11-20.  I look at players who are likely to be undrafted in rounds 21-23.
• Use DVS in a few mock drafts before your actual drafts.  This way, you'll be used to it if you use it on draft day.  It can really make a difference if you know how to use it properly.
• Even if you declare the SGP concept to be bunk, DVS still functions as a live scoring spreadsheet.

And, before anyone asks - it is unlikely I'll be able to accommodate other options into DVS.  This means I probably won't be able to develop one that includes the things that IPDS does [league types, sizes, roster configurations, etc].  I wish I had the time, but all of my drafts are either Friday or Saturday, and I need some time to do personal preparation for those.  Also, since the University of Houston is on spring break, I'm looking to throw some baseballs and work on my breaking ball this week.

An extra special thanks goes out to Jordan Lyall and his great work.  Without it, DVS wouldn't have happened.  Without any further delay, click here to download DVS.  Please pass along the word about Nine Bo Jacksons - prepping for the draft is only the beginning the great season I'm planning on writing about here.

Comments

[Derek]

I have been experimenting with a similar concept this season. My tool is a little different, as it simulates the remainder of the draft after each pick based on team positional needs and my program's estimation of the user's willingness to deviate from default rankings, as well as how I score those deviations. Then the adjustments to a given player's category scores are made based on the estimated final rankings of the complete draft (and a handful of other things, such as the probability that another player will take them before my next pick).

The category adjustments are also made in such a way that I can start leveraging the data even in early rounds (no AJ Pierzynski factor). The one draft I did using this tool came out amazingly well ... although I think the strategy was more advantageous in this league given that there were 26 scoring categories (13x13 rotisserie). Not sure if it would have been quite so helpful in a normal 5x5, but as always, more information is good. My tool allows sorting based on a number of different scoring algorithms, and I find each of them useful at different times in the draft.

[Jordan Lyall]

Nice Job, Ike!

I love what you've done with it. I hope many fantasy team owners get a lot out of the tool.

- Jordan

[Brent]

I'm not exactly an excel ninja..is there a way to adjust this more teams?

[Dan]

this is spectacular. very well done.
one suggestion: maybe for next year, you guys can figure out a way to set up a separate sheet where we can easily track the picks of the other teams by position (for instance, so i know if everyone the guy picking after me has his OF filled up already)

[Jermaine]

If this tool came out too late to use during the draft (Like for me) It's a good way to gauge your leagues standings preseason once you plug all the info in. Nice.

[Vino Sentry]

I've tweaked the sheet a bit to support less than 12 teams. You can download it from http://bigtendrunks.com/DVSv21.zip.

I put a couple of notes on the main tab outlining how to use and a disclaimer related to the new ADP column.

[dave]

Ike,
Does the DVS include Tim's 3rd round of projections (which are apart of IPDS)? This stuff is great!

Vino,
What does the 7th column (ADPRD10) represent? The ADP for a 10 team league? ADPRD and ADPRD10 match when you set it to 11 teams. I'm just trying to figure this out.

[Bob]

You said:
"You can unhide the columns to reveal each players statistics and categorical SGP data. I set it to six columns in order to keep all of the vital information on my small screen at the same time. You can expand and contract it as you'd like."
How do you do this???
I would like to see each players stats.

[argystokes]

Vino (and maybe Ike): I've downloaded Vino's version, and it appears to reward high ERAs and WHIPs; that is, the team with the best ERA earns just 1 standings point.

[Ike Warner]

Brent:

There isn't an easy way to adjust it to more teams. Logistically, it is possible, but only with time that I don't have right now. Lo siento.

[Ike Warner]

Dan:

That's a great idea. I have a whole lot of ideas to build a drafting program for next season, and yours is definitely part of it. After all - I have to have something to do in November!

[Ike Warner]

Vino:

Well done, sir.

[Ike Warner]

Dave:

The current version of DVS doesn't include Tim's recent projections. I plan on updating both DVS and IPDS with them today.

[Ike Warner]

Bob:

Depending on what version of Excel you're on - when you're on the Draft Sheet tab, hit Ctrl+A, then go to Format Cells, Unhide Columns. This should expand all the hidden cells. If you type "Unhide Columns" into Excel's help file, it will give you a step-by-step on this.

[Ike Warner]

argystokes:

I can't speak for Vino's version, but I want to point you back to the post's warnings on SGP points - they won't make a whole lot of sense until later in the draft.

"You'll also notice that the starting pitchers have a SGP value of 22 when you start up. Why? Because even Johan Santana's 3.07 ERA and 1.06 WHIP is going to be "worse" than the 0.00's that the rest of the league has."

[walkoffblast]

"It's a good way to gauge your leagues standings preseason once you plug all the info in. Nice."

Thats how I ended up using it as well. It also becomes an interesting tool to judge preseason trades with since you can get a feel for how it helps you in the categories you might be trying to improve.

[argystokes]

Ike - The problem appears to be specific to Vino's version. Vino, you'll probably want to check it out (see my comment above).

[Justin]

With regard to argystokes' problem on Vino's worksheet, you need to make one small change to the formulas that rank WHIP and ERA. For each cell, change the last "1" in the formula to a "0". This reverses the order it ranks the ERA and WHIP numbers.

[Vino Sentry]

ADPRD10 is a bad column name. It basically attempts to apply a linear multiplier to the ADP in the original column based upon the number of teams in your league. It worked for 10 and 8. I didn't test odd numbered league sizes less than 12 - I did it at work during my lunch.

[Ben Ice]

Unfortunately this tool is pretty useless for those of us that do a lot of monoleagues or in mixed leagues with larger # of teams. It's a nice tool, but really only applicable to the less complex leagues.

Not bashing it, as I was very interested in using the tool. But my mixed leagues are typically 14-16 teamers, and my 12 teamers are AL or NL only.

[Ike Warner]

Ben:

I really wanted to make it with the flexible options that IPDS had, it just wasn't possible given the time frame I had to work with. Something to do for next year, I suppose.

[Max]

I am entering my draft data from a few days ago now to play with this sheet. When I got to round 8 the sheet started to remove team names from the drop down list. First I assumed maybe you can only enter someone's starting lineup, but I got to a guy who only has 2 pitchers on his roster and his name is not available for the pitcher he drafted.

Is there an error or am I using it wrong somehow? This problem began soon after I had started playing around with the sorting options in the SGP column, but I didn't do anything that should have broken the sheet. Team name page is filled in and still correct.

Using WinXP Excel 03 SP2.

Thanks

[Max]

I found a work-around for my issue in case anyone else gets the same problem. If i right click the 'Drafted By' box and choose the 'pick from drop down list' option it refreshes the list and the missing team names re-appear.

[Ike Warner]

Max:

Sorry to hear about the issues you've had. I haven't had such errors - but thanks for posting the solution for anyone who has.

[Kevin K.]

thanks for that fix !! not a stat geek or even an excel person and got to the point where I was entering in a teams draft choices for outfield and the team had disappeared from the drop down list yet was still on the first sheet. panic just about took hold until I remembered seeing something here about a solution