AltTab

Project Zomboid reviews: who actually stops enjoying it

Project Zomboid News 8 min read

Data basis: 800 English-language Project Zomboid reviews, pulled from Steam’s public appreviews endpoint (appid 108600) on 2026-08-27, sorted most recent first. They span 22–27 August 2026. Playtime figures are the reviewer’s own recorded hours as Steam reports them at review time. This is a snapshot of one week, not the game’s lifetime score — Steam’s all-time figure sits far higher, at roughly 191,600 positive against 13,300 negative.

Steam shows you one number: the percentage of reviews that are positive. What it does not show you is who is unhappy. A player who refunds after ninety minutes and a player who quits after 1,400 hours both register as one thumb down.

Splitting the same 800 reviews by recorded playtime turns that single number into a curve, and the curve has a shape worth looking at.

Satisfaction peaks at 50–200 hours

Positive review rate by playtime Project Zomboid — 800 recent English reviews, 22–27 August 2026 0-2h 9% (n=11) 2-10h 76% (n=68) 10-50h 88% (n=230) 50-200h 92% (n=285) 200-1000h 89% (n=177) 1000h+ 79% (n=29) Satisfaction peaks at 50–200 hours and falls again above 1,000. n=11 in the 0–2h band is too small to lean on.

Read up the bands and approval climbs exactly the way you would expect: barely-played reviews are harsh, and every additional band is kinder than the last — 76% at 2–10 hours, 88% at 10–50, peaking at 92% in the 50–200 hour band.

Then it turns. The 200–1,000 hour band comes in at 89%, and past 1,000 hours it falls to 79% — lower than the 10–50 hour band. The people who have played most are not the people who rate it highest.

Two cautions before anyone builds a theory on that. The 1,000-hour band is 29 reviews, and the 0–2 hour band is 11 — both small enough that a handful of reviews moves the number several points. And a review written at 1,400 hours is not a verdict on 1,400 hours; it is usually a verdict on the last twenty.

What the shape does support is a narrower claim: the discontent in this week’s reviews is not coming from people who bounced off the game. It is coming from people well past the point of having learned it.

Mods are the most common complaint — by a distance

What negative reviews talk about Share of reviews mentioning each topic, negative (n=101) vs positive (n=699) mods 26% neg · 9% pos B42 bugs 23% neg · 4% pos UI / controls 15% neg · 2% pos performance 14% neg · 2% pos multiplayer 14% neg · 2% pos learning curve 4% neg · 2% pos Keyword matching, not human coding — a review can match more than one topic. Method stated in the article.

Sorting the 101 negative reviews by topic, mods lead. 26% of negative reviews mention mods, the workshop, or modded saves — roughly one in four. Build 42 bugs follow at 23%, then UI and controls, performance, and multiplayer, clustered around 14–15%.

The comparison column matters more than the ranking. Mods appear in 9% of positive reviews too, so some of that is enthusiasm rather than complaint. But every other topic on the list is three to seven times more common in negative reviews than positive ones — those are complaints, not mentions.

Method, stated plainly: these are keyword matches, not human-coded categories. A review saying “the mod scene is incredible” and one saying “my mods are all broken” both land in the mods bucket. One review can match several topics. The counts are reproducible from the same endpoint; the interpretation is ours.

The timing question this raises, and cannot answer yet

The sample window straddles a patch. 42.20.4 went out at 20:12 on 26 August and, as the 42.20 hotfix timeline covers, it removed the Lua loadstring and loadstream methods — a change the announcement itself warns will break mods that used them.

Splitting the sample at that timestamp:

WindowReviewsNegativeMods mentioned in negatives
Before 42.20.476190 (11.8%)20%
After 42.20.43911 (28.2%)45%

The negative rate more than doubles and mod-mentions among negatives climb from a fifth to nearly half. It is tempting to draw a line straight from the patch to the reviews.

We are not going to. The post-patch window is 39 reviews covering a few hours. At that size the negative rate has a wide error bar, and only one review in the entire 800 mentions loadstring or loadstream by name. Reviewers who are annoyed that their mods stopped working generally do not know which method was removed.

The honest statement is that the shift is visible and the direction matches, and that a week of further reviews would settle it. This page will be revisited then rather than guessing now.

What this is useful for

  • Deciding whether to buy right now: the 2–10 hour band — the closest proxy for a new player’s first sitting — sits at 76% positive in this week’s sample. That is the number that applies to you, not the all-time 93%.
  • Running a modded save: mods are this week’s leading complaint topic, and a patch three days ago deliberately broke a modding method. Check your mod list before updating, not after.
  • Coming back after a long break: much of what the 1,000-hour band is reacting to is Build 42 changing rules they had internalised — which Build 41 knowledge expired lists what actually stopped being true.
  • Reading review scores generally: a single percentage hides which players are unhappy. Here, the unhappy ones are disproportionately the most experienced, and that is the opposite of what a headline score implies.

What could not be confirmed

  • Whether 42.20.4 caused the post-patch negative spike. Direction matches, sample is 39.
  • Whether unhappy long-playtime reviewers are reacting to Build 42 specifically or to accumulated hours. Review text was not coded for that, and playtime does not record when those hours were played.
  • Non-English reviews. The pull was English-only; a different language mix could shift every figure here.
  • How the sample compares to the same week a month ago. Steam’s endpoint returns recent reviews, not historical windows, so no before-and-after baseline was available.

Sources

Related posts