Fantasy NBA Player Rater

I know this is a football blog, but I'm a big NBA fan as well and with the season starting in a few days I've been getting ready for some fantasy drafts and came up with a simple model to rate players, and a nifty graphic to show which players might be worth picking up and which are currently being drafted too highly.

Quick disclaimer, this model uses ESPN's own projected stats which can be found here, so the model will only be as good as those projections are. If you disagree with those stats then let me know, but I've got faith that ESPN's analyst are some of the best out there (that provide data for free atleast). The projected stats were taken on the 23rd October.

Total Standardised Score

My model is built on a system of standardised score, where players performance in each scoring category is calculated as a value relative to the average. If a player is exactly average in one category then their score is 0, a positive score means they are above average, a negative score means below average. The score for each category is then summed up into one measure: Total Standardised Score. The advantage of this is it allows you to see the overall impact a player will have on your fantasy team. For example, a player might offer you a lot of assists, but if he's costing you elsewhere then is he beneficial to the team? Likewise, a player like Howard might cost you in FT%, but do his numbers elsewhere make up for this?

You can find more info on this method here, ESPN's system seems slightly more complex than mine, but I hope the use of mine is clearer.

Average Draft Position vs Total Standardised Score

Below is the graphic I wanted to post. Along the bottom you have the average position players are being drafted at in the ESPN game, I've selected the Top 100 as beyond the 10th round you'll be looking to flesh out your team and pick players that compliment those you've already drafted. Along the y-axis we have Total Standardised Score, the higher the better.

Learn About Tableau

Each data point is a player, mouse over and you can see who it is. The size of the data point is the standard deviation across all categories. Smaller circles imply production across all categories while large circles imply excelling in some categories but falling behind in others. For example, Dwight Howard has a large circle because of his high FG%, Blocks and Rebounds, but below average FT%, 3s and Assists. Paul George has a small circle because he contributes well in all categories.

The bands show the different rounds, assuming a 10 team league.

Finally, the graph is interactive, you can click on the positions below to highlight each in term.

Comments and Position Tips

The first thing to note is the negative relationship, so better players are going earlier, good start. More interesting is when we split players by position and see how the score deteriorates over the rounds.

Most strikingly, the standardised score for PFs does not differ much throughout the first 10 rounds, Kevin Love is not rated much higher than Derrick Favors who is going 7 rounds later. The shallowness of the best fit line shows how there is not a great dropoff at the PF positon.

On the flip side, the SF position does deteriorate through the rounds, there's a top 6 (James, Durant, Anthony, George, Batum, Smith) then a huge drop off.

There's a good spread of Centres and Point Guards throughout the Top 100, so take them as you can. Though again there's an elite group of 4 PGs (Curry, Paul, Irving, Williams) who would all make good first round picks. The depth at PG in the NBA means you can still pick up a good PG as late as late as the 7th in someone like Bledsoe.

SG is an odd position, there's Harden, who should be drafted Top 5 if you can, then a huge drop off. Wade is an option, but his playing time will surely be down, and Bryant is an unknown. After that otherwise I'd let this position slide to the 7th and get someone like Mayo or Oladipo if your happy to take a rookie.

Overall I'd target one of the top SF or PG first, then look for a C, then worry about the PF and SG spots last. Though of course it depends on your pick, if you can take Harden 3rd then Alridge at 18th, do it.

Player Tips

Just some quick tips on who is being overlooked so far and who is worth leaving till later rounds.

Pickup:

• Kemba Walker - gives you everything you need at PG and is probably going late because of the team he plays for rather than his stats
• Paul Millsap - doesn't hurt you in any category and excels in Steals and Assists for a big man
• Jeff Teague - solid pickup as your first bench option
• Jimmy Butler - quickly rising up through the ranks in the last few days and with good reason. Great production across the board.

Pass On:

• Marc Gasol - great player, but probably overvalued in the 1st. Howard and the two Al's are better options.
• Brook Lopez - sits well below the trendline for centres, blocks aren't enough here.
• JR Smith - will end up hurting you with poor FG% and little defensive contribution.

Risk:

• Derrick Rose - former MVP in the 2nd round? Big question over fitness but could be a steal.

Wrap Up

I hope this post has been interesting to someone, I'd love to hear some feedback on my first NBA post. If anyone's interested I'll post the spreadsheet up for people to download.