This morning my Facebook friends are up in arms about a proposed re-valuation of Scrabble tiles in line with letter frequency in modern language usage:

Scrabble: Should letter values change?

All of them (admittedly they are mostly more casual gamers than me!) are united in the opinion that any normalisation of the tile values would be tampering with, and probably even breaking, something that isn't broken.

My own viewpoint is that the only real consequence of such a change would be to reduce the swinginess inherent in the game from, e.g. drawing and scoring big points from the Z. I want to know whether it would be fair of me to tell my friends, "well, by all means argue for keeping the tile values as they are, but be aware that what you're arguing for is keeping the ratio of skill to luck in the game weighted towards luck".

Of course perhaps it's me that's missing the point, and there would be unforeseen negative consequences to gameplay and fun from reassigning the letters with "fairer" values. Can anyone think of any?

  • I would think that changing the values of the tiles would need to go hand-in-hand with changing the actual quantity/ratio of the tiles available to ensure balance.
    – goldPseudo
    Jan 15, 2013 at 15:54

4 Answers 4


The changing dictionary, especially with regards to the 2-letter words (notably ZA and QI), certainly influences what a "good" scoring of the tiles would be.

A situation to consider is A and U. If you have 3 A tiles, you can play AA, usually for 8+ points because you can put them in front of a wide variety of high-value consonants (BDFHKMPZ), and quickly rebalance your rack. If you have 3 U tiles... good luck. Maybe you can play ULU for 3 if you're lucky. It's probably not correct to have A and U both be worth 1 point.

One thing that surprised me in their proposed scoring was the reduction in the value of C. Like V, C is not in any 2-letter words in the North American dictionary, and is thus much harder to play for good value in closed-off boards. That might be balanced by its prominence in 7-letter words, but I'm skeptical.

It sounds like this evaluation was generated from a computer which knows 100% of legal words, which can definitely shift the balance around quite a bit. Dealing with a V is much easier once you know VAV and VAW, for example, but only enthusiasts know those words. A "human-tuned" set of tile scores is going to be much harder to reason about.

One thing I think is worth calling out is status quo bias. If X, which can easily be dumped and often scores huge plays because of its flexibility in XI/XU/AX/OX/EX, were worth 5 points and it was proposed that it go to 8, experienced Scrabble players would probably say "That's idiotic, you're just giving a windfall to whoever draws it".


The article says that the changing dictionary has changed the values of letters, and this seems like a very good point. For example, the existence of a single word, ZA, listed as entering the language in the mid-late 60s, makes the Z much easier to play. (I wish I had an old dictionary to figure out when the even-more-important QI entered English dictionaries!)

However, there already exists a "Scrabble-inspired" game with different tile values, of course: Words With Friends. (That game seems to have made their letter value choices based on different guidelines than the article; for instance, its G is still more valuable than an M.) Players who are good at Scrabble are still good at Words with Friends; I would not expect to see any differences in win percentages below the master level of play. Ultimately, your luck in your tiles' value is largely outweighed by your luck in drawing tiles that go well together; a good blend of consonants and vowels, for instance, and plenty of the powerful S (which would be the best tile to draw even if it were worth 0).

  • And let's not forget the Scrabble blank - the best tile to draw, and it IS worth 0! Jan 15, 2013 at 14:33
  • Both ZA and QI along with 3,300 or so other words were added in the Fourth Edition of the Official Scrabble Players Dictionary in 2006.
    – ghoppe
    Jan 15, 2013 at 18:37
  • @thesunneversets nitpick: because the blank is worth 0, it's not always the best tile to draw. It's the tile that gives you the best chance to play, but there will (nearly) always be a tile that could have given you more points. ;)
    – ghoppe
    Jan 15, 2013 at 18:41
  • 2
    I'd be willing to play under the condition that my rack always consisted of 7 blanks. I think I'd win. Jan 16, 2013 at 18:38

Changing values on tiles will not change the final goal. During the game you will still have to place letters with big values to get lots of points and win the game. The letters will change but not the fact that if they get a high value, placing them will help you to win the game.

Now in English those letters seems to be 'Z', 'Q', 'X' and 'J' (referring to Wikipedia, I don't play with the English version) but with the new values it will be other letters. That's all.


Firstly, Words With Friends changed these letters scores:

L, U, N: move from 1 to 2 G from 2 to 3 H and Y from 4 to 3 B, C, M and P from 3 to 4 V from 4 to 5 J from 8 to 10

The rest have stayed the same. Although K and V both score 5 now there is only one K tile and still two V tiles. There are now more than 4 S tiles that regular scrabble gives you. There seem to be a lot of H and Y tiles.

A bingo has gone down from 50 bonus to only 35 which I quite like because it means there is less emphasis on playing one. And I reckon that, and not the scores of the tiles, is the biggest difference between the game and potential tactics.

And of course the bonus square positions have moved.

A lot of scrabble players seem to put emphasis on 2-letter words and frequency of letters appearing in them, which makes 'C' one of the worst tiles as it doesn't appear in any (along with V). It is useful to learn the 3 letter words with 'V' and I used to have a rule to get rid of it quickly even if only on single-letter. 'C' is useful in longer words if you have an open board and bingo opportunities.

Compare that with the 'F' that appears in a lot of short words but isn't that good for longer ones.

Now, does the changing of the scores of the tiles have a major impact in the game? I would say, not noticeably. The positioning of the bonus squares and the lower bingo bonus may have put less emphasis on aiming for a bingo and more on better placement.

How it would have a negative impact would only really be if the game became more dependent on being lucky enough to pick up high-scoring letters rather than low-scoring ones. Or, if the elevation of 'L' and 'N' here to 2 points meant the relative weakness of having a 'Q' for example was not compensated by its higher scoring value when you place it.

I think there are variations of the game that would put less emphasis on being lucky enough to hold a balanced rack (vowels and consonants), and that there were 2 bags, a vowel back and a consonant bag, and the 2 blank tiles put aside. When you draw, you can choose which bag to use, and in a 2-player game each player may choose when to pick up the blank among their letters. As you draw you see what you get before seeing what to draw next.

Then luck of the letter draw would become less of an issue.

And it's no use saying "good players know how to balance their rack". How do I control what I pick up if I bingo? How do I stop myself drawing vowel after vowel or consonant after consonant no matter how I play them?

  • Did you mean to post this as an answer to boardgames.stackexchange.com/questions/8988/…?
    – jwodder
    Oct 23, 2014 at 22:04
  • I answered that question too. The question here is really subjective anyway, it might be better worded "what tactical difference would it make".
    – CashCow
    Nov 5, 2014 at 11:35

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .