Back to index of Scrabble pages by Donald Sauter.
Please visit the Scrabble For Word Lovers introduction page.

 
When I meet someone, and it comes out we're both Scrabble players, the next thing I usually hear is, "How many points do you score in a game?" Then I find myself explaining once again that points per game is not a very telling statistic under the best of circumstances, and in the case of someone like me who plays a mix of two-, three- and four-person games, is completely worthless.

The Scrabble statistic that always does the job is average Points Per Turn (PPT), and why the Scrabble world has not embraced it ranks up with the great mysteries of the universe.

Just as your batting average in baseball tells how likely you are to make a hit in any trip to the plate, your PPT in Scrabble tells how many points you're likely to score on any turn.

As good as the analogy is, there's a profound difference. In baseball, batting average tells only part of the story. One player may bat .320 hitting mostly singles; another may have the same average with a goodly number of doubles and home runs mixed in. Then again, a player may have a stellar batting average, but be a very weak fielder or runner.

In Scrabble, your Points Per Turn average is everything.

To understand why I say this, let's return to the more familiar Points Per Game statistic. Besides being a very coarse statistic, it's hardly interesting because the total points that can be scored is more or less limited regardless of how good you get.

Think about it: a Scrabble game always uses up the same tiles on the same board with the same pattern of premium squares. That yielded about 620 points per game back in the days of the first OSPD. (It would be somewhat greater now with the addition of QI, QAT, ZA, etc. I don't have data to work with for play with the OSPD4.) Two novice players may take 21 turns apiece to get there, while better players may take only 16. But they end up at the same spot. Heck, if two chickens could be trained to kick Scrabble tiles onto a Scrabble board, they would score in the same ballpark. (Be advised: sometimes I exaggerate ever so slightly to make a point and wake the reader up.)

Yes, there can be exceptional games where every power tile gets counted six times, but such games owe far more to the luck of the draw than developed skill. I'm assuming even those "novice" players above know how to look around the board for bright, colorful squares when they get a J, Q, X, Z, etc.

Expert players score significantly higher, but that's almost wholly due to their higher bingo production. Take away those 50-point bonuses and their combined final scores would look much like ours, somewere in the 600s. Put the other way around, if you did detect a small upward creep in your average Points Per Game, it would be mainly due to increased bingo production. And that's something which would be more fun and useful to track separately. (Do you know your MTBB -- Mean Turns Between Bingos?)

Scrabble is about grabbing up those limited, available points in as big handfuls as possible -- and that's exactly what Points Per Turn measures. Just add up all the points you score and divide by the turns taken. Ignore the adjustment for the leftover tiles, but you must count your 0-point turns! If you traded tiles, you did so for the sake of future points. If you got burned playing a bum word, well, you struck out!

We've talked about batting average and PPT. There is another fantastic difference. A batting average only makes sense in the context of the "league" in which you play. A 12-year-old slugger batting .564 in his Little League probably wouldn't do so well against Major League pitching.

Your PPT in Scrabble, on the other hand, is almost rock-steady no matter what level of competition you find yourself in. After all, what difference does it make whether the board you are stewing over was cobbled together by an expert or by a duffer? Ironically, if you are a recreational player with a 19.6 PPT, say, you would probably see it rise a hair against experts, since their longer words open up more scoring opportunities for you.

Since there are so many turns in a game, a very meaningful PPT can be calculated in just a few games. It doesn't depend on games always being one-on-one. It would not be affected by a game being cut short, or games using more or less than 100 tiles. (In my club, we scoop a fresh set of about 100 tiles for each game from a 300-tile set.)

PPT shows you how your point production improves over time. It gives the most meaningful basis of comparison between yourself and all other Scrabble players. It provides a benchmark to gauge the play you just made. ("Uh oh, I only scored 30 points there. Now I gotta make 36 on the next one!")

So PPT is useful, you say, but that doesn't make it "everything"; it only measures your offensive abilities. My rebuttal is that there is no appreciable defense element to Scrabble play, that Scrabble is very nearly a 100% offensive competition, that no one wins Scrabble games by trying to keep his opponent from scoring. As I said above, your mission is to grab up points while you can. It is folly to think you can get through a game of Scrabble keeping your opponent off the premium squares all the while. If you make a play just to tie things up, you are slitting your own throat as much as his.

I have found that few, if any, Scrabble players agree with me on this point. The best I can do is offer an invitation. Stop on by for a game. All of your attention will be on defense. You will play solely to keep me from scoring points. We won't even tally up or record your plays. Your mission will be one thing only -- to see how far below my current PPT you can drag me. Good luck.

I personally would like to see Points Per Turn, which is simplicity itself, replace the tournament rating system, which involves an inscrutable formula known to a select few (I hope). Unless I misremember, even the author of the Scrabble bestseller "Word Freak", a book about the author's quest for a certain rating, never learned for himself how it was done. I know that every now and then the voodoo formula is rejiggered so that everybody's rating jumps from one arbitrary number to another.

If the Scrabble rating ranks players in a different order than PPT, one would be hard-pressed to explain how that could be. You will generally beat players with a lower PPT, and lose to players with a higher PPT. If you score more points per turn than anyone else, then you are the best Scrabble player on earth. (Congratulations!)

Now, then, what's your PPT?

In case the Points Per Turn statistic sounds so obvious that you can hardly believe it's not used by Scrabble players, consider a few observations culled at the time of writing this, May 2009.

A Google search on - scrabble "points per turn" - turned up only 62 hits (after Google weeds out "very similar" pages.) Several of these pages were, in fact, duplicates. Eleven of the hits were my own writings, either from my own web site, or pages that drew from my posts to rec.games.board.

With one interesting exception, the PPT mentions of any relevance were with respect to just a single game, or perhaps a set of games, never a running statistic indicating a specific player's skill.

The interesting exception was in Hasbro's Scrabble FAQ:

My wife scores 400-450 points per game. Would she be considered an expert?

Not necessarily. At SCRABBLE clubs and tournaments experts average between 330-450 points per game. However, the competition is probably much stiffer than your wife encounters, and she may have to adjust her thinking to adapt to the typical 25 minute time limit per person per game when using a chess clock. A better measure of skill is determining the "average points per turn" score. If your wife averages 30 or more points per turn, not counting tile exchanges, then she may very well be a SCRABBLE expert. The very top SCRABBLE players average 35 or more points per turn, not counting exchanges. We suggest she visit and play at a SCRABBLE club or tournament and see how she fares.

This was actually lifted from the book "Everything Scrabble", by Joe Edley (2001), page 227, and reworked slightly. Hasbro courteously worked in "my wife" for "I".

So this is the only mention I've ever come across, by someone other than me, of average Points Per Turn as a useful statistic in Scrabble. I don't recall any mention of it in Word Freak, for instance.

I can't swear that I was the first person on earth to think it up, but let the record show that I calculated the PPT statistic for everyone in the Bowie Scrabble Club (Maryland) right from the start in mid-1985. And much of what I've written about PPT above was part of my original Scrabble page, which went up on the web as soon as I got internet access in 1997.

Getting back to the FAQ, what's this business about "not counting exchanges"? Although a Scrabble player scores no points on a turn in which he exchanges his tiles, he chose to do it in the service of maximizing his point production for the game. It counts.

And I'm guessing the "35 or more points per turn" for top Scrabble players, was somewhat inflated for 2001. In the 34 Table 1 games of the 2015 North American Scrabble Championship, including 5 championship playoff games, the average PPT was about 33.1. And remember that this is after the Official Word List was beefed up in 2005 with thousands of new Scrabble "game pieces", most notably, the game-changing QI and ZA.

Contact Donald Sauter: send an email; view guestbook; sign guestbook.
Back to Donald Sauter's main page.
Back to the top of this page.

Helpful keywords not in the main text: ospd = official scrabble players dictionary.