5-Letter Words for Scrabble High Score — Suffix Scoring Economics & Morphology Analysis
Not all five-letter word endings are equal in Scrabble. -CK words average 13.29 points — 4.13 above the dataset baseline. -ER words average 8.16 — a full point below it. The difference is not which words happen to be in each group. It is tile mathematics embedded in the suffix structure itself.
This analysis examines why certain five-letter endings score systematically higher, using computed suffix data from the 12,478-word verified dataset. The question is not just what scores well — it is why.
Source: 12,478-word verified five-letter word set
Scoring: Standard Scrabble tile values, base score only (no board multipliers)
Baseline: Dataset average 9.17 pts across all 12,478 words
Analysis: Suffix groups filtered from full dataset; avg, differential, no-repeat ratio computed per group
Guide: Full scoring overview at Scrabble Scoring Guide
-CK is the highest-scoring five-letter suffix class at 13.29 avg (+4.13 above baseline, 90th percentile). -ER is below average at 8.16 despite its 434-word size. Z+ED interactions average 16.24 — nearly double the -ED class average. Scoring advantage in Scrabble endings comes from tile values embedded in the suffix, not the words themselves. → Find plays from your rack at the Unscrambler
Suffix Scoring Distribution — Computed From 12,478 Words
Twelve common five-letter suffixes compared against the 9.17-point dataset baseline. The differential column shows how much each suffix class exceeds or falls short of average.
| Ending | Words | Avg Score | vs Baseline | Percentile | No-Repeat % | Top Word |
|---|---|---|---|---|---|---|
| -CK | 75 | 13.29 | 90th | 72.0% | QUACK (20 pts) | |
| -LY | 130 | 10.34 | 73rd | 49.2% | JOWLY (18 pts) | |
| -TY | 136 | 10.28 | 73rd | 64.7% | ZESTY (17 pts) | |
| -RY | 93 | 10.06 | 73rd | 51.6% | QUERY (17 pts) | |
| -ED | 516 | 9.47 | 62nd | 71.5% | HAZED (18 pts) | |
| -NG | 87 | 8.48 | 49th | 90.8% | KYANG (13 pts) | |
| -ER | 434 | 8.16 | 49th | 66.1% | HAZER (17 pts) | |
| -ND | 49 | 7.94 | 32nd | 91.8% | CHYND (14 pts) | |
| -AL | 163 | 7.79 | 32nd | 64.4% | ZYGAL (18 pts) | |
| -LE | 145 | 7.79 | 32nd | 76.6% | AIZLE (14 pts) | |
| -ST | 110 | 7.40 | 32nd | 80.9% | QUIST (14 pts) | |
| -SS | 35 | 7.23 | 32nd | 0.0% | QUASS (14 pts) |
The range is striking. -CK (13.29 avg) outperforms -SS (7.23 avg) by more than 6 points per word on average — not because of the specific words in each group, but because of the tile mathematics embedded in each ending. -SS also has 0% no-repeat by definition: SS always repeats S.
Why -CK Dominates — Structural Tile Mathematics
-CK's scoring advantage is not a coincidence of which words happen to end in -CK. It is a consequence of tile arithmetic: K scores 5 points, C scores 3 points — the two-letter ending alone contributes 8 base points before any other tile is counted. A five-letter -CK word with three common 1-point tiles (like SLACK: S+L+A=3 pts) still scores 11 — above average.
This is structural scoring. Any valid -CK word starts with an 8-point foundation. The question is what the first three letters add. Words like QUACK (20 pts) and QUICK (20 pts) push higher because Q(10) compounds the advantage; words like SLACK (11) and BLACK (12) use common consonants at the front.
The Y-Suffix Effect — Why -LY, -TY, -RY Cluster Together
Three of the four above-average suffix classes end in Y. This is not a pattern — it is a structural cause. Y scores 4 points. Any word ending in -LY, -TY, or -RY carries a 4-point Y tile in the final position, automatically lifting the word's base above what the other letters alone would achieve.
-LY words carry Y(4) + L(1) = 5 points from the ending. The 49.2% no-repeat rate is notably lower than comparable suffix classes — many -LY words repeat vowels (BADLY, ODDLY, EARLY) or consonants (DULLY, FULLY). This is worth noting for rack strategy: -LY words more often burn two positions on the same letter.
-TY carries Y(4) + T(1) = 5 points. Similar total to -LY, but with better no-repeat characteristics — 64.7% carry five distinct letters vs 49.2% for -LY. -TY words also show stronger rack-balance profiles because T appears commonly in support positions without dominating.
Why -ER Falls Below Average — The Hidden Scoring Trap
-ER is the largest suffix class at 434 words — and among the lowest-scoring. E(1) + R(1) = 2 points from the ending, the lowest tile contribution of any suffix examined. Most -ER words build their score from the first three letters alone, and those letters are typically common consonants and vowels.
This creates a structural trap for Scrabble players: -ER words feel productive (large word pool, familiar vocabulary) but consistently score below average. WATER (8 pts), TIGER (8 pts), LINER (5 pts) are all below the 9.17 baseline.
High-Value Letter Interactions — Where Causation Gets Specific
The most analytically interesting finding is not which suffix classes score highest overall — it is which letter-plus-suffix combinations create scoring outliers within otherwise average groups.
| Interaction | Words | Avg Score | vs Class Average | Why |
|---|---|---|---|---|
| Z + -ED | 17 | 16.24 | +6.77 vs -ED avg | Z(10) as primary driver; -ED ending uses low-value tiles, amplifying Z's relative contribution |
| X + -ER | 12 | 13.33 | +5.17 vs -ER avg | X(8) rescues a structurally weak suffix; BOXER, HEXER, FIXER, VEXER |
| K + -CK | 75 | 13.29 | Already the class (K is the ending) | K IS the CK ending's primary driver; the suffix and the letter are the same mechanism |
| K + -ER | 36 | 10.36 | +2.20 vs -ER avg | K(5) lifts the otherwise weak -ER class; BAKER, MAKER, POKER |
| Y + -LY | 130 | 10.34 | All -LY words have Y (Y is the ending) | Y IS the LY scoring driver; same mechanism as K+CK |
The Z+ED interaction is the most striking finding. -ED words average 9.47 points. Z+ED words average 16.24 — a 6.77-point premium from a single letter substitution. The -ED ending's low tile values (E=1, D=2) make it a near-perfect amplifier for Z: the 10-point tile contributes a larger fraction of the total score than it would in a higher-baseline suffix.
No-Repeat Patterns — Where Morphology Meets Rack Efficiency
Suffix classes vary significantly in their no-repeat rates — and the variation is not random. It reflects the linguistic structure of English morphology rather than vocabulary selection.
| Ending | No-Repeat % | Why |
|---|---|---|
| -ND | 91.8% | N and D rarely repeat in English vocabulary; -ND naturally avoids duplication |
| -NG | 90.8% | Same structural pattern — N and G don't self-duplicate in English word formation |
| -ST | 80.9% | S and T are common but avoid each other in repeated positions |
| -LE | 76.6% | L and E appear frequently but rarely double in five-letter LE endings |
| -LY | 49.2% | Many -LY adverbs contain doubled vowels (ODDLY, DULLY, FULLY) or consonants |
| -SS | 0.0% | SS always repeats S — structural impossibility of no-repeat in this class |
-ND and -NG lead at 91.8% and 90.8% no-repeat respectively — the two highest no-repeat rates among all suffix classes examined. Both feature consonant pairings (N+D, N+G) that rarely appear doubled in English vocabulary, making them structurally productive for rack efficiency despite scoring below the dataset average.
The -SS class at 0% is a definitional constraint, not an anomaly: every word ending in -SS contains a repeated S. These 35 words are structurally excluded from the 8,013-word no-repeat pool regardless of their other characteristics.