How to Nail Your Third Wordle Guess (When You Have 10 Letters Tested)

The Puzzle That Used All Six Guesses

I'm not proud of this story. Last month I started a Wordle with SLATE, got the A in the right spot and confirmed the T wasn't in the word. My second guess, CORNY, gave me the O in position 4 and confirmed R wasn't in the word either. By guess 3, I had _A_O_ with five letters confirmed absent.

I guessed VALOR. Wrong. Then MAYOR. Wrong. Then FAVOR. Wrong. I used all six guesses and failed.

Afterward I looked at the Wordle answer archive. The word was MANOR. If I'd spent guess 3 on an elimination play instead of guessing blindly, I'd have had it by guess 5 at worst. That game is the reason I started thinking seriously about what the third guess actually needs to do.

What You Actually Know by Guess 3

With two non-overlapping 5-letter starting words, you've tested exactly 10 unique letters. If you're using something like SLATE + CORNY (a popular two-word combination), you've covered S, L, A, T, E, C, O, R, N, Y. That's 10 of the 26 letters — including most of the high-frequency vowels and consonants.

By now you should know:

• Which of your 10 tested letters are in the word (yellow or green)
• Where the confirmed letters do or don't belong (greens tell you position, yellows tell you the letter exists but not where)
• That 10 specific letters aren't in the word at all (gray tiles)
• Roughly how many valid words still match all your constraints

The problem is that last point. Most players never actually calculate how many words remain. They feel confident, make a guess, and are surprised when they're wrong. The number of remaining possibilities is the single most important factor in deciding what to do at guess 3.

Reading Your Color Pattern

Before making guess 3, take 15 seconds to read your board. Your color pattern tells a story.

Lots of Green Tiles (3+)

If you've confirmed three or more positions by guess 3, you're in great shape. With three greens and one yellow, the number of remaining possibilities is usually very small (often fewer than 5 words). Guess the answer. The risk of wasting another elimination guess outweighs the benefit of narrowing down one more letter.

Two Greens and a Yellow or Two

This is where it gets interesting. Two greens plus yellow information constrains the word significantly, but depending on the pattern, you might still have 6-12 options left. Before guessing, think about whether those remaining options share a pattern. If they do, you may be looking at a cluster trap (more on that in a moment).

One Green or Mostly Yellow

You know your letters are in the word but haven't pinned their positions. This often means more remaining options than you think. Treat this as a likely elimination-guess situation unless one candidate word clearly fits all your yellow constraints perfectly.

Mostly Gray

Heavy gray means your starting words didn't produce many hits. The good news: you've eliminated 10 letters, which is genuinely useful. The remaining word must use only the 16 untested letters plus whatever yellows you have. But you probably still have 30-100+ valid candidates. An elimination guess is almost certainly the right call here. Don't guess blindly.

The Two Paths at Guess 3

There's really only one question to ask yourself at guess 3: How many valid words still match everything I know?

If the answer is 1-3 words, guess one of them. You have a 33-100% chance of getting it right, and even if you're wrong, you'll nail it on guess 4.

If the answer is 4 or more words, you need to think carefully. Especially if those 4+ words share a pattern, because that's the cluster trap situation.

The Wordle solver can tell you exactly how many valid answers remain after entering your current clues. I started doing this mid-game (yes, just to count, not to cheat) and it completely changed how I approach guess 3. Knowing there are 8 remaining words instead of 3 makes the elimination guess decision easy.

The Cluster Trap: Wordle's Sneakiest Pattern

A cluster trap is when the same letter pattern fits many different words, and all of them are common enough to be Wordle answers. The most brutal examples:

The _IGHT trap: LIGHT, MIGHT, NIGHT, RIGHT, SIGHT, TIGHT, FIGHT, BIGHT, DIGHT, WIGHT. If you've confirmed the _IGHT pattern by guess 3, you could have up to 10 valid candidates. Guessing randomly gives you at best a 1-in-10 chance per guess. With 3 guesses left, that's genuinely risky.

The _OUND trap: BOUND, FOUND, HOUND, MOUND, POUND, ROUND, SOUND, WOUND. Eight valid candidates, all common English words.

The _ATCH trap: BATCH, CATCH, HATCH, LATCH, MATCH, PATCH, WATCH. Seven candidates, all easy words Wordle would absolutely use.

I hit the _IGHT trap twice in the same month once. First time I guessed through it and failed. Second time I recognized the pattern immediately and played an elimination word. Got it in 5. Still not great, but a clear improvement over failing entirely.

Related traps worth knowing: _ANGE (MANGE, RANGE, GANGE), _ASTE (BASTE, HASTE, PASTE, TASTE, WASTE), and _ORD (CORD, FORD, WORD, LORD, GORД). Any time you have 3+ letters confirmed in positions 2-5 with position 1 unknown, ask yourself how many words fit.

My honest take: the _IGHT trap is single-handedly responsible for more broken Wordle streaks than any other pattern. It looks solved — you have four letters confirmed — but it isn't. The trap is precisely that it feels solved. Players who recognize it improve immediately. Players who don't keep calling it bad luck.

How to Build a Good Elimination Guess

When you decide you need an elimination guess, the goal is simple: test as many of the likely first-letter candidates as possible in a single 5-letter word. This requires a bit of quick mental work.

Take the _IGHT scenario. You've already used SLATE + CORNY, so S, L, A, T, E, C, O, R, N, Y are off the table. The potential first letters for _IGHT words include F, M, D, B, H, W, P, G. A good elimination guess would be a valid 5-letter word containing as many of F, M, D, B, H, W, P, G as possible.

You won't find a word with all 8, but even covering 3-4 of them cuts your remaining candidates from 8+ down to 2-3. At that point, guess 4 almost certainly ends the game.

The same logic applies to any cluster trap. List the untested first-letter options. Find a valid 5-letter word that tests as many of them as possible. Play it on guess 3. Read the result. Guess the answer on guess 4.

When Going for the Answer Is the Right Call

Elimination guesses are powerful, but they're not always the right call. Skip the elimination and go straight for the answer at guess 3 when:

Three or more green tiles. If you've confirmed 3+ positions, your remaining candidates are usually very limited. Even if there are 4-5 possibilities, guessing one of them is usually better than a risky elimination play.

Two greens plus strong yellows. If your yellows heavily constrain where the remaining letters can go, you may only have 1-2 valid options left even with just two greens. Think it through quickly before committing to an elimination guess.

You can clearly see the answer. Sometimes by guess 3, one word obviously fits all your constraints perfectly. If it feels obvious and no other word comes to mind, trust it. Don't overthink. The most common mistake here is second-guessing an obvious answer.

You're playing hard mode. In hard mode, you must use all confirmed letters in every subsequent guess. This limits your elimination options significantly. If you're playing hard mode, see the hard mode strategy guide for how to handle cluster traps under those constraints. It's a different problem entirely.

Three Worked Examples

Example 1: Strong Position, Guess the Answer

SLATE → (A in position 3, green; T in position 4, green) CORNY → (O in position 1, green)

By guess 3, you have O _ A T _ with no confirmed 5th letter. This pattern fits very few common words. Think: OATHS? No, you need the A in position 3. ORACY? Doesn't fit. With O_AT_ you're actually looking at very few candidates. QUAFF doesn't work. It's likely OPATR or something unusual. In this case, if OVATE comes to mind, play it. Strong position.

Example 2: Cluster Trap, Elimination Guess

SLATE → (nothing green or yellow) CORNY → (nothing green or yellow)

Ten letters tested, none of them in the word. You've eliminated S, L, A, T, E, C, O, R, N, Y. The word uses only untested letters: B, D, F, G, H, I, J, K, M, P, Q, U, V, W, X, Z plus the 4 vowels I and U. Remaining candidates: VIVID, WHIFF, BUMPH, GUILD, QUIFF, SKIMP... you have dozens of possibilities. This is a definite elimination-guess situation. Test the most common untested letters first.

Example 3: The Classic Two-Green-and-Yellow Situation

SLATE → (S in position 1, green; E in position 5, green) CORNY → (nothing new)

You have S _ _ _ E confirmed. That fits: SINCE, SWINE, STOVE (wait, T was gray from SLATE)... Actually you know T and L and A are absent. So S _ _ _ E with no L, A, T, C, O, R, N, Y. Words: SWEDE, SWIPE, SPIKE, SPIRE, SHIVE, SHIRE... still quite a few. Count them before guessing. If it's more than 5, lean toward elimination. If it's 3 or fewer, guess one. Our second guess guide covers how to set yourself up for cleaner information coming into this exact scenario.

What the Data Says

According to the WordleBot FAQ from the New York Times, the optimal play at any point in the game depends on how many possible answers remain. When many answers remain, WordleBot consistently recommends elimination guesses (which it calls "guesses that maximize information") over direct answer attempts. The threshold where guessing an answer becomes mathematically better than eliminating is around 2-3 remaining options.

Research published on Towards Data Science analyzing Wordle strategies found that players who use elimination guesses strategically rather than always guessing answers average significantly fewer total guesses per game. The improvement is especially pronounced for players who start in the 4-5 guess range, exactly because they're encountering cluster traps and not recognizing them. (I'll be upfront: the exact percentage improvement varies across studies and depends heavily on which version of the answer list was analyzed. The directional finding holds, but I wouldn't cite a specific number.)

The common Wordle letter patterns guide goes into the statistical frequency of different letter positions. Understanding which patterns repeat is the fastest way to recognize a cluster trap before it costs you guesses. And checking words ending in a specific pattern (like -IGHT or -OUND) helps you see instantly how many candidates you're actually dealing with.

The Five-Second Gut Check

Before every guess 3, ask yourself one question: can I name three or more different words that fit everything I know right now?

If yes, consider an elimination guess. If no (one or two words come to mind), guess the answer.

That's it. You don't need to enumerate all possibilities in your head. The simple gut check catches most cluster trap situations and prevents the worst outcomes. I run this check every single time now, and my guess 3 decision-making is dramatically better for it.

Check out 5-letter words and words ending in E to see how many candidates even a simple constraint can leave open. It's eye-opening, and it'll calibrate your instincts for when you're actually in trouble vs. when you think you are.