Ranked #9 on the Russian sports Top Grossing chart with only one review: "Туфта" — how an AI football prediction app got on the chart

On Nov 2, an AI football prediction app called Scora charted at #9 in the Russian App Store Sports Top Paid chart, ahead of names like UFC and KHL. Yet it had only 2 ratings in Russia, averaging 2.5 stars, and the only review with text was one word: "Туфта" (scam). It didn't break out. It charted because small-country charts have low thresholds and the weekly subscription is priced high. This piece explains how that kind of chart entry happens, and who actually makes money in the "AI football predictions" niche.

Nov 2, #9 in Russia

That day the top of Russia's App Store Sports Top Paid chart included PeakVisor (peak identification), Oceanic+ (dive computer), UFC, PPPoker (poker) and KHL (hockey league). Scora sat at #9, next to two utility apps. Two days earlier, on Oct 31, it was at #41.

It showed up in other countries the same week: #15 on the Tanzania Sports Top Grossing chart (11-04), #22 in Colombia (11-06), #62 in Indonesia, #77 in New Zealand, #112 in the UAE. On the surface, it looks like a new app selling well in a dozen countries at once.

Look closer and it doesn't fit. Per iTunes data on Nov 6: 0 ratings in the US store, 0 in Colombia, 2 in Russia. Across all countries, ratings rose by only 1 over 10 days. On Google Play it has 1K+ downloads. The Russian review is one star, with "Туфта" as both title and body.

What it is

Scora uses a "neural network" to predict football matches: who wins, the score, the number of goals, plus live scores, fixtures and stats. The homepage of scora.me reads "over 80% accuracy" and "stop guessing, start winning," and lists per-category accuracy: match result 84%, over/under 78%, and exact score 64%. Hitting the exact score six times out of ten is a number nobody who watches football would believe (inference: a normal rate is one in ten or so). The site also claims 50K+ downloads and 4.9 stars, which don't match the store numbers.

Pricing: weekly subscription $5.99, monthly $14.99, yearly $89.99. In Russia it's ₽499 per week. Age rating 17+. The site FAQ states "we are not a betting operator, we only provide information." In plain terms, it sells to people who want to bet on football, packaged as an information tool.

开发者叫 SF COMPANY,TOV,TOV 是乌克兰的有限公司形式(推断)。这个开发者账号 2020 年前后注册,以前发过做短视频模板的 App,现在只挂 Scora 一个。商店名按国家翻成当地语言:美区叫「AI Soccer Predictions」,俄区「ИИ Прогнозы на Футбол」,哥伦比亚「Pronósticos Fútbol IA」,台湾「AI足球預測與賽事數據分析」,可 App 本身只标了英文。苹果给的首发日期是 2026 年 7 月 13 日。

One order can move you a hundred places

The pattern is the same in every country: it jumps up one day and falls back within a few days.

App Store Sports Top Grossing charts by country, 10-28 to 11-06.

The reason is simple. In these countries' Sports Top Grossing charts, apps in the middle and lower ranks typically make very little revenue in a day. For a $5.99-a-week subscription, one extra payer or renewal can move the ranking a long way. Flat ratings with erratic rank swings usually mean one or two orders (inference, unverified: monthly revenue probably in the tens to hundreds of dollars).

I couldn't find where its users come from. Searching its name on TikTok turns up only a skincare brand with the same name. On X, there have been zero posts in nearly 90 days, and no media outlet has covered it. The site has a standalone landing page with only an App Store button, which looks like it's for ad campaigns, but I couldn't find any of its ads. The only pattern visible: it's named in each country's language, with names stuffed with words like "AI," "prediction" and "score," so store search picks it up in small markets like Angola, Belarus, Venezuela and Ghana.

How crowded this niche is

Scora is not an outlier. Since Oct 28, 656 apps with names containing words like predict, прогноз, pronóstico, tips or betting have charted on App Store sports charts across countries. 197 of them launched this year, and 98 launched after June. The App Store's "You Might Also Like" links them into a chain that feeds traffic among themselves, and each one gets very little of it.

The ones that actually got big are these:

Google Play figures as of about 11-06.

Beyond these there's another layer: sites like aipredict.io and nerdytips give predictions for free and live on search traffic; Sofascore and FotMob offer free scores, stats and fan polls; and if you want to know who wins tonight, you can just ask ChatGPT.

Next to these, Scora's selling points (picking bets by expected value, a player-contribution model, 40 bet types, 80% accuracy) are nearly all offered by similar apps. It has no public track record, no content, and no volume push.

Can this category be built

The money is real. Selling a high-priced weekly subscription to people who bet, packaged as an information app and passing review as 17+, is the standard playbook across the whole category. 180Score and TipsPro both show it can reach hundreds of thousands or millions of installs.

But nearly two hundred have entered this year alone, on nearly identical templates, and users can't tell them apart. The ones that made it have two things newcomers can rarely get quickly: time and volume (180Score spent two and a half years and spread to nearly a hundred countries), and a clear, verifiable pitch (ScoreGPT's "multi-model voting + public track record"). There's another risk too: Apple reviews apps with gambling overtones strictly. Scora's App ID was created in September 2025, but it didn't go live until July 2026, a ten-month gap that may reflect review delays (inference).

If it were you

In one line: there's nothing about Scora worth copying. The AI football prediction route does make money, but it's already packed, and a newcomer will most likely be number 657 in line.

• Who it suits: people already producing sports content with an existing audience (a football WeChat account, a TikTok football channel, a fan community for one league), or teams with paid acquisition experience in a smaller-language market who are willing to fund a weekly subscription with ads.
• Who it doesn't suit: indie developers hoping to get organic traction from store search. Hundreds of template apps compete for the same keywords, and these apps face uncertain review, refunds and bad reviews.
• If you do build it, how: don't copy Scora, copy ScoreGPT. Publish every prediction and its result on the site so accuracy can be checked, which is what this category most lacks. Use per-match pre-game analysis pages on the website to capture search traffic. Pick a league or language the big apps ignore (for example, a domestic league in one country, in its local language).
• How to acquire users: short videos and pre-match analysis articles around match days, with public track-record screenshots. Calculate paid acquisition carefully, because weekly renewal rates decide whether you break even.
• When to move: mid-season for European leagues, before next summer. That's when demand for these apps peaks each year.
• Signs you've misread it: if you've publicly tracked results for two months and accuracy is no better than random guessing, or if the payback period on paid acquisition never comes down, stop putting money into this.

One last point, which applies to anyone scanning charts for opportunities: a rank on a small country's Top Grossing chart, if ratings don't rise and it jumps a hundred places in a day, is most likely just one or two orders. Check whether ratings grew first, then check the rank.