Yesterday, I talked about how splitting players up into tiers before a fantasy draft can be a useful exercise. Today we’ll look at some of the problems that tiers can create.
Consider the 3B values the CHONE projections give for a standard fantasy league:
| Alex Rodriguez |
$43 |
| David Wright |
$41 |
| Aramis Ramirez |
$19 |
| Garrett Atkins |
$19 |
| Chipper Jones |
$12 |
| Ryan Zimmerman |
$12 |
| Chris Davis |
$10 |
| Kevin Youkilis |
$10 |
| Adrian Beltre |
$9 |
| Edwin Encarnacion |
$9 |
| Chone Figgins |
$8 |
| Aubrey Huff |
$7 |
| Troy Glaus |
$7 |
| Ian Stewart |
$6 |
| Evan Longoria |
$5 |
| Carlos Guillen |
$5 |
| Mark Reynolds |
$4 |
| Jorge Cantu |
$4 |
| Hank Blalock |
$2 |
| Alex Gordon |
$1 |
| Kevin Kouzmanoff |
$1 |
| Melvin Mora |
- |
| Mike Lowell |
- |
We’ve decided that there’s clearly a top tier of A-Rod and Wright — no one else is valued at even half of their prices. The second tier is also easily demarcated by a $7 drop after the next two players, Ramirez and Atkins.
But do you see any tiers after Atkins? I’m seeing a pretty steady decline from $12 to $1.
Let’s say we decide to split the remaining players into roughly equal-sized groups, drawing the line between Troy Glaus and Ian Stewart. That gives nine players from $7-12 and eight from $1-6.
But does it make sense to say that a $6 player is closer to a $1 player than he is to a $7 player? That would be prefering a $5 difference to a $1 gap, which doesn’t seem logical to me.
This is the problem with tiers: They work fine for about the upper 20-25% of players, where you can find clear separations between groups of players. But for most of the draft, there just aren’t big gaps in value between players. In these situations, I’d argue that attempting to separate players based on arbitrary dollar value cutoffs does more harm than good.
Tiers of Common Traits
So what can we do for players in the mid-to-late rounds of a draft? One idea is to group players by similar traits instead of strictly relying on dollar values. For example, with the 3B above, we might form a fantasy tier of Solid Yet Possibly Declining Veterans with Chipper, Beltre, Glaus, Guillen, Mora, and Lowell. We might form a Young, High Risk/Reward Tier with Davis, Stewart, and Longoria.
With tiers formed by common characteristics, you can get some insight that you don’t get with value tiers. If your early picks at other positions were geared towards players with lots of upside (and downside), maybe you should draft the best available 3B in the Veteran Tier. A consistent performer like Beltre can bring some balance and stability to your team.
If you have played it safe in the early rounds, maybe now is the time to take some risks. I highly doubt Longoria will still available, but Ian Stewart is an intriguing player if he gets some playing time. Why not roll the dice and and see if you can grab a bargain?
Stat Category Tiers
Maybe we go a slightly different route and form tiers based on common statistical contributions, while relaxing the positional requirements. We make a tier called Late Speed Options that has Figgins, Pierre, and Taveras. There’s a Good Batting Average Tier with Polanco, Loney, Kendrick, and Helton.
Suppose you get near the end of the draft and your team is looking light on AVG. You have an OF spot open and a MI spot available, so you can look in the Good Batting Average Tier for the best player left that you can put in one of those two spots. That player might not be the highest player on your overall draft board, but he could still be the best choice for your team.
Do you see any other ways to divide up players into meaningful tiers?
Subscribe