Most players who lose consistently are not losing to bad cards. They are losing a point here and two there, in the same handful of places, every single game. Here they are in order of how much damage they actually do.
A note on the numbers. Where we give a figure below it is per game, against an opponent of similar strength, and it is an estimate from ordinary play rather than a computed result — unlike the figures in the hand odds, which are exhaustive. Treat them as an ordering, not as measurements.
Cost: 2–4 points a game, and the largest single item on this list.
The discard is not one decision, it is two different decisions that happen to look identical, and which one you are making depends entirely on who deals. Into your own crib you are throwing cards that will score for you, so 5s, pairs and cards that combine are gifts to yourself. Into your opponent's crib exactly the same cards are gifts to them.
Players who have learned "keep the best hand" as a single rule throw the same two cards in both cases and hand their opponent a four-point crib every second deal. Before you choose, say whose crib it is out loud. It is a trivial habit and it is worth more than any other single change a beginner can make. There is a proper treatment in the discard guide.
Cost: 2–4 points a game.
Almost everybody spots 9 + 6 and K + 5. Far fewer reliably spot 4 + 4 + 7, or A + 2 + 3 + 9, and those are worth exactly the same two points. Every distinct group of cards totalling fifteen scores, regardless of how many cards it uses.
The fix is procedural, not clever. Count fifteens in a fixed order every time: all the pairs of cards first, then all the triples, then any group of four. Saying "pairs, triples, fours" under your breath is enough. The hand calculator lists every fifteen it finds on its own line, which makes it an unusually good drill for this particular blind spot — count a hand yourself, then see which line you missed.
Cost: 1–3 points a game, and occasionally a whole game.
The starter is part of every hand and the crib. It belongs to nobody and it counts for everybody. Beginners look at the four cards in their hand, count those, and stop — which reliably loses whatever the fifth card was contributing.
The tell is that your counts come out suspiciously round and low. If you are regularly counting 4s and 6s and your opponent is regularly counting 8s and 10s from similar-looking cards, this is almost certainly why.
Cost: 1–2 points a game.
The 5 is the most valuable card in cribbage because sixteen cards in the deck are worth ten, so any 5 has sixteen partners waiting to make it a fifteen. A 5 in your opponent's crib is very rarely worth less than two points to them and often worth four.
The rule of thumb is close to absolute: never put a 5 into a crib that is not yours unless the alternative is genuinely dreadful. Players break it to keep a marginally prettier hand, and it is nearly always the wrong trade — you are giving away a near-certain 2 to gain a speculative 1 or 2.
Cost: about 1 point a game each, and they are free.
Two separate points, easily confused, both routinely dropped:
Heels is the more commonly lost of the two, because it has to be taken at a moment when nobody is thinking about scoring yet. If you deal, look at the starter before you look at anything else.
Cost: 1–2 points a game, and the errors are large when they happen.
A hand of 3 4 4 5 holds two separate runs of three — one using each 4 — and scores 6 for runs, not 3. Add the pair and the fifteen and the hand is worth considerably more than it first appears. The same logic makes a double run of four worth 8, and a triple run worth 9.
The mistake is to see "a run" as one thing that scores once. It is better to think of it as: find the longest run, then count how many different ways you can make it. Two ways is double, three is triple. Missing the second half of a double run costs three points at a stroke, which is why this ranks above several more frequent errors.
Cost: around 1 point a game.
Leading a 5 into the play offers your opponent a fifteen with any of the sixteen ten-cards, which is the most likely single response in the deck. It is the one lead almost every experienced player avoids automatically, and one that beginners make constantly because a 5 feels like a card to get rid of.
A 4 is the conventional lead — it makes fifteen only with a Jack, Queen, King, Ten or an 11-combination, and it sets up your own run possibilities. More on this in the pegging guide.
Cost: around 1 point a game.
Taking the count to 21 invites the ten-cards, and there are sixteen of them: your opponent reaches 31 for 2 far more often than chance suggests. The same applies to leaving the count at 5, which invites the same sixteen cards to make fifteen.
Counting to 11 or 26 is the mirror image and works in your favour — both are positions where a ten-card cannot be played safely. Thinking about what a ten-card would do to the count, every time, is most of pegging defence.
Cost: variable, occasionally catastrophic.
You move your rear peg, but you count from your front peg. Counting from the rear peg instead is the classic board error and it silently awards or robs you of your previous score. Because the board is the record, nobody may notice for several deals, at which point the game is unrecoverable.
The two-peg system exists precisely to make this visible — the gap between your pegs is your last score, on display. See the board explained.
Cost: hard to quantify, and it decides close games.
The right discard on the first deal is not the right discard when you are on 105 and your opponent is on 118. Late in a game, "how many points do I need" beats "which hand is biggest": if you need four to win, a certain four beats a probable seven, and if your opponent needs three, denying them the crib matters more than building your own hand.
Most players never make this switch at all and play deal one and deal ten identically. Simply asking "what does each of us need?" before the last two or three discards is enough to get most of the benefit — and it is the one item on this list that separates good players from merely accurate ones.
Six of the ten above are counting errors, and counting errors respond well to repetition. Open the hand calculator, press Deal me a hand, count it aloud before looking, then check the breakdown. Twenty hands takes about ten minutes.
What matters is not the score you get but which line you missed. Nearly everybody is wrong in the same three places every time — a three-card fifteen, the second half of a double run, or nobs — and once you know which is yours, it stops happening within a week.
Then put it to work. Play free against the computer — it counts every hand, so you can check yourself deal by deal.
Play Cribbage Now