How well each casino app is built
Every app here does roughly the same job. They are not equally good pieces of software, and the differences are measurable before you install anything: how big the download is, how long it takes, how old a phone it still runs on, and whether anyone has touched it lately.
The comparison
Ordered by build score. Install time assumes a 20 Mbps connection and covers the download only — not unpacking, not first launch, not the account setup after it. The two scores are kept apart deliberately: see below for why an app with one rating does not get to beat one with eight hundred.
| App | Build | Users | Size | Install | Needs | Updated |
|---|---|---|---|---|---|---|
| Odogwu Bet | 10.0 | too few ratings | 57.8 MB | 24 s | iOS 13.0+ | 16 d ago |
| 1xBet | 9.3 | 5.3 1,599 ratings | 269.6 MB | 1 min 53 s | iOS 15.0+ | 18 d ago |
| Numba | 9.3 | too few ratings | 94.1 MB | 39 s | iOS 14.0+ | 32 d ago |
| ZARbet | 9.3 | 9.3 814 ratings | 34.7 MB | 15 s | iOS 15.1+ | 15 d ago |
| LulaBet | 8.7 | too few ratings | 110.7 MB | 46 s | iOS 13.4+ | 123 d ago |
| 1win | 8.3 | too few ratings | 1.9 MB (wrapper) | 1 s | iOS 16.0+ | 44 d ago |
| Odibets | 8.3 | 8.2 240 ratings | 50.1 MB | 21 s | iOS 16.6+ | 32 d ago |
| Wasafibet | 8.3 | 8.7 349 ratings | 39.2 MB | 16 s | iOS 14.3+ | 158 d ago |
| Surebet247 | 7.7 | too few ratings | 73.7 MB | 31 s | iOS 13.0+ | 14 d ago |
| Gwalabet | 7.0 | too few ratings | 3.6 MB (wrapper) | 2 s | iOS 15.0+ | 228 d ago |
| LEON | 7.0 | too few ratings | 93.5 MB | 39 s | iOS 15.0+ | 18 d ago |
The size difference is the one that changes how the app feels
Of the apps here, 2 are under 10 MB. An app that small contains no game artwork, no sounds and no interface to speak of. It is a browser window with an icon: the operator's website, loaded fresh over the network every time you open it.
That is not dishonest and it is not always worse — it installs in a second and updates without the App Store. But it decides what "slow" will mean for you. A wrapper is exactly as quick as your connection at that moment, on every screen, every time; it has nothing cached to fall back on; and on a weak signal it shows you a blank page where a native build would show you the interface and wait for the data.
Where the two scores disagree
Only 4 of these apps have enough ratings for the user column to mean anything, and among those the two numbers come apart hardest on 1xBet: 9.3 on the build record against 5.3 from 1,599 people who installed it. ZARbet scores 9.3 and 9.3 — the record and the reception agree.
That gap is the honest limit of everything else on this page. A recent update, a low iOS floor and an iPad build tell you the app is maintained and broadly compatible. They do not tell you it is pleasant to use, and here is a case where a large number of people who did use it disagreed. Read the build score as what the developer has done, and the user score as what came of it.
How the score is calculated
The score is the mean of the axes below. Each is derived from one field in Apple's store data by a fixed rule, so two apps with the same data get the same score and re-running the sweep moves every number at once.
- Kept up to date
- Days between the current version's release and the date we checked. Under a month scores 10, a year or more scores 1. An abandoned betting app is the one that breaks on the next iOS release and stays broken.
- Runs on older phones
- The declared minimum iOS. iOS 13 or lower scores 10, iOS 17 or higher scores 4 — graded from the reader's side, because a high floor is a choice the developer makes and a phone the reader cannot use.
- Built for more than one screen
- Whether the listing carries iPad screenshots and declares a universal binary. Both scores 10, neither scores 3. An iPhone-only build on an iPad runs in a letterboxed window.
- What people who installed it say — reported, not averaged in
- The App Store rating on the same 0–10 scale, and the only figure here reflecting the app running on real phones. It is kept out of the build score on purpose. We built it in as a fourth axis first, averaging over whatever each app had — and the app with a single rating came out top of this table, ahead of one rated 4.72 by 814 people. Averaging across a missing axis does not represent the uncertainty, it deletes it, and it rewards the app nobody has used. Below 50 ratings the column says so instead of showing a number.
- An axis we built and then deleted
- Apple's data appears to carry a rating for the shipping version separately from the lifetime average — which would have shown whether the current build is landing worse than the app's history, the nearest thing to a performance trend readable from outside. It is an alias. Every app returns an identical figure and an identical vote count in both, one of them on 3,708 ratings. Scored, it would have given every app the same mark while looking like a measurement, so it is not here. We mention it because the number is public and someone else will publish a chart of it.
Download size and install time are reported but deliberately not scored. Small is better for the download and worse for everything after it, and we are not going to pick a curve that decides which of those matters more to you.
What this cannot tell you
Frame rate. Launch time. Whether it drops your bet slip on a bad connection. Whether its push notifications are useful reminders or marketing you will turn off in a week — Apple exposes none of that, and no amount of reading the store page produces it.
Getting it would take an Apple Account registered to a storefront that carries the app, a real device, and someone timing it. That is a real piece of work rather than an impossible one, and if we do it the numbers will appear here, dated, next to the ones above — clearly separated, so you can always tell which figures came from a measurement and which from Apple's metadata. Until then this page tells you what the store record supports and stops.
All figures from Apple's own store data, captured 9 August 2026. How we verify.