Author here. I scored every word in the ENABLE open word list (172,835 words) under both the Words With Friends and the standard Scrabble letter-value tables, to see how far apart the two games land on the same word.
The part I didn't expect was how one-sided the gap is. A word can score up to +11 more in WWF — agammaglobulinemic is 41 in WWF vs 30 in Scrabble — but the biggest edge the other way is only −3 (dysrhythmia, 23 vs 26). It's structural: only H and Y are worth more in Scrabble than in WWF, while ten letters (B, C, G, J, L, M, N, P, U, V) cost the same or more in WWF. So there's no high-value Scrabble-favoring letter for a word to lean on — every −3 word is stacked with H's and Y's. Across the full list only 8.02% of words (13,862) score identically; the mean absolute gap is 2.528 points.
One honesty note: this uses ENABLE, the free/open list a lot of open-source word games use — not the copyrighted TWL (NASPA) or CSW (Collins) tournament lists, and it isn't affiliated with either game. The tile values are the two games' own published tables.
Author here. I scored every word in the ENABLE open word list (172,835 words) under both the Words With Friends and the standard Scrabble letter-value tables, to see how far apart the two games land on the same word.
The part I didn't expect was how one-sided the gap is. A word can score up to +11 more in WWF — agammaglobulinemic is 41 in WWF vs 30 in Scrabble — but the biggest edge the other way is only −3 (dysrhythmia, 23 vs 26). It's structural: only H and Y are worth more in Scrabble than in WWF, while ten letters (B, C, G, J, L, M, N, P, U, V) cost the same or more in WWF. So there's no high-value Scrabble-favoring letter for a word to lean on — every −3 word is stacked with H's and Y's. Across the full list only 8.02% of words (13,862) score identically; the mean absolute gap is 2.528 points.
One honesty note: this uses ENABLE, the free/open list a lot of open-source word games use — not the copyrighted TWL (NASPA) or CSW (Collins) tournament lists, and it isn't affiliated with either game. The tile values are the two games' own published tables.
Every number is reproducible — the script pulls the same word list the site ships and computes all of it: https://gist.github.com/lucian-devops/21fff2526147b5dcc88af0... . Happy to get into the scoring method.