
A football probability app built solo in Ashgabat cracked Uzbekistan's sports Top Grossing chart at #3
What it is A web-plus-app football probability tool: for each match it lays the model's probability side by side with the probability implied by the odds, publishes an archive of daily predictions, runs no ads, and charges by subscription ($3/week, $5/month, $38.4/year).
Why it took off Starting September it used small Meta ad budgets to test the website, then from around November 5 switched to driving App downloads; it defaulted to Russian, added Uzbek, Turkish and other languages, and opened Mir card and SBP payments on the web for Russia, breaking into the sports Top Grossing charts in the low-barrier Uzbek and Russian stores.
CHow you could follow it Stop building another football predictor. Move its structure to another category or project: pick a language region you know, where Western products only ship in English and local payments are blocked; build the website first to collect money, then add the app; for Meta ads, make a separate creative set per language, test the website first, then switch to App download ads.
Est. monthly revenue (Google Play) under $1KWe estimated this from its rank on country top-grossing charts and real install counts (it isn't on the grossing chart, so confidence is lower), calibrated against real revenue; typical error is 2–3x. Read it as a tier, not an exact number.
- Data (updated daily) · first launch 2026-10-21 · 11-19 installs 1,066
Android · Total installs worldwide Data 2026-11-11 → 2026-11-19 (9 days, 9 observations). Google's detail page only refreshes install counts every other day, so a flat day doesn't mean nobody installed.
On November 17 and 18, VARStats ranked #3 in the App Store Uzbekistan Sports Top Grossing chart, ahead of big names like Setanta Sports, Tennis TV and UFC. It was built alone by Merdan Allaberdiyev, an indie developer in Ashgabat, the capital of Turkmenistan; the iOS version launched October 28, and by November 19 it had only 1,066 Google Play installs. Volume is small, and the "football prediction" category is extremely crowded. What's worth a look is the playbook: target languages Western products can't be bothered with, catch money others can't collect, and use small-budget Meta ads to buy the first users. This piece covers how much of that playbook is learnable and why this category isn't worth chasing.
What it sells
Open varstats.com and the first screen reads: "The probability of every result, over/under, corner and card, next to the probability the odds imply. You see not just how we read the match, but where we and the market disagree." Each match gets its own page, with the model's probability, the odds-implied probability with the margin removed, and the gap between them, row by row. It covers 100 leagues, including the Uzbek Super League and the Iraqi league.
It deliberately avoids saying "guaranteed win." Merdan's own description on the indie launch site Fazier reads: "Football sites either give you raw data or give you a 'prediction', and never tell you how many times they were right." So VARStats writes its "Today's picks" into a public archive before each day's kickoffs and never edits them afterward. On the homepage as of November 19 it states: in the last 30 days, 246 of 346 matches hit (71%). On November 2 he published a post on the Russian sports site Sports.ru titled "I checked 281,621 football predictions: when the model said 70%, it happened 72.6% of the time." The website and app run no ads and carry no betting affiliate links, which is rare in this category.
Monetization is by subscription: $3 a week, $5 a month, $38.40 a year (original price $48, promo code valid through November 30), with a 3-day free trial and no card required. The website and app share one account and can be paid for on either. The website also has an entry reading "Pay from Russia: Mir card / SBP" and accepts cryptocurrency.
How it came out of nowhere: three unremarkable moves
First, small Meta ad budgets: test the website, then push the app. The domain varstats.com was registered in September 14, and by September 18 the first Instagram post and first ad were already live, pointing to the website. In mid-October several ads ran for only one to three days, and one had accumulated just 3 hours of spend, clearly creative testing. A Turkish creative read: "Intuition can fool you, numbers speak… 18+ · probabilities are not guarantees." From November 5 and 6, Russian and Turkish "Download VARStats free" ads began running, and another new one went up on November 17 (Meta Ad Library, 11-19).
The timing lines up: the App download ads went live on November 5, and the same day the iOS version appeared at #31 on the Russian sports Top Grossing chart; on November 7 it climbed to #27 on the Uzbek sports free chart; on November 11 it showed up at #186 on the Google Play Turkey sports free chart. How much was spent is not visible from outside.
Second, languages first, Russian at the top. The website defaults to Russian, with Turkish, Turkmen, Uzbek, Kazakh and Azerbaijani, 16 languages in total. The sitemap has about 37k pages, with every match and team page carrying versions in 16 languages, with titles like "Albacete - Sabadell: bepul prognoz" (Uzbek: free prediction). The robots.txt contains rules written specifically for Yandex, suggesting he is targeting Russian-language search. These pages haven't brought much traffic yet: the site doesn't appear in the global top 1M, and October Chrome user data places it only in the top 500k in Turkey.
Third, catching money others can't collect. Since 2022, Russian users can't pay on Google Play, and on Apple only mobile-carrier billing remains. VARStats added Mir cards and SBP (Russia's local fast-payment system) on the web, plus cryptocurrency. An English-language prediction app wouldn't build for this market; a Russian-speaking Central Asian developer just did it on the side.
What ranking #3 does and doesn't tell you
| Where | Rank | Date |
|---|---|---|
| App Store Uzbekistan Sports Top Grossing | 4 → 11 → 6 → 8 → 3 → 3 → 9 | 11-13 to 11-19 |
| App Store Russia Sports Top Grossing | 31 → 112 → 14 → 44 | 11-05, 11-12, 11-16, 11-19 |
| App Store Greece Sports Top Grossing | 36–82 | From ¥11-15 |
| Google Play Turkey Sports Top Free | 186 → 131 → 164 | 11-11, 11-15, 11-19 |
| Google Play global installs | 412 → 618 → 838 → 1,066 | 11-11, 11-14, 11-17, 11-19 |
What you can read: someone is paying, and on iOS. What you can't read: how much was earned. The Uzbek and Russian sports Top Grossing charts have low barriers, so a few subscriptions can push a ranking up (inference). Play has no ratings at all, and iOS has only 2 ratings each in Uzbekistan and Russia, with around 70 to 80 installs a week; that's the scale.
Who it's competing with
Almost everyone. "Football probabilities + subscription" is one of the most crowded categories in 2026:
- "AI prediction" apps on the same charts. As of November 19, more than twenty Turkish-language "AI tahmin" (AI prediction) apps sat in the top 200 of the Google Play Turkey sports free chart: TipsPro AI at #30, Scouter at #39, GoScore AI at #40. On the iOS Russia sports Top Grossing chart, FootyStats ranked #6, and AiXTIP from a Turkish developer ranked #8 (17.7k Play installs); on the Uzbek chart, GG AI Football Predictions ranked #7, right next to VARStats, pitched as "AI guaranteed win."
- Heavy-spending big players. 11XScore AI Football (3.08M Play installs, up 7k in 53 days), 180Score AI (2.94M).
- Established sites. Forebet has been doing math-based predictions for 2014 years, has a Russian version, and 3.58M Play installs; BetMines has 7.16M installs.
- Already on users' phones. Sofascore (108M Play installs), Flashscore, FotMob; they show no probabilities, but scores and stats are all there.
VARStats is a latecomer to this wave, not its originator. It has 0 discussion on X, a TikTok account with 1 followers, and two Sports.ru blog posts with only 5 and 7 likes respectively, and nobody has copied it. Its "public calibration" gives the ads something to say (the English creative line: "Our accuracy is public"), but rivals using "AI guaranteed win" still landed on the same charts, and there's no sign users are coming for the "transparency."
If it were you
One-line verdict: C. What's worth learning is the playbook (pick a language region you know and Western products ignore, pair it with local-usable payments, website first, small-budget ads to test creatives); what's not worth following is the category. Football prediction is already crowded with "AI guaranteed win" clones, and VARStats itself only proves "someone pays in a low-barrier region's charts," not that a real income can be made.
Who it suits: one or two people whose native language is Russian, Turkic, or another language big tech hasn't localized for and whose speakers will pay, who can build data products and run their own Meta ads. Better to apply this playbook to a different category than to build another football predictor.
Who it doesn't suit: anyone aiming at English, Portuguese or Turkish markets (dozens or hundreds of competitors, with big players gaining tens of thousands of installs a week); anyone who can't stomach the review risk and payment-channel risk tied to betting-adjacent content; anyone hoping to grow on a "we're more transparent" differentiator without buying traffic.
How to build it: if you must do sports probabilities, don't go head-to-head in football. Try what it doesn't do: basketball, tennis, esports, or a single country's local league, served in the local language in as much detail as possible. On product, copy its approach: one page per match, odds-implied probability shown beside the model, predictions archived before kickoff and never edited, no ads. Let the paid website and the app share an account, and offer local payment methods on the web. Most valuable is to take this structure to a different category: find a market where Western products only ship in English and local payments are blocked, collect money on the web first, then launch the app.
How to get users: at its pace, Meta ads go to the website first, with a separate creative set per language; some run only a few hours before being stopped, and once something works, it's switched to App download ads. Programmatic SEO on multilingual match pages lays the groundwork, but don't expect it to bring volume in the short term: after two months VARStats' search traffic is still small. Set the goal first at reaching the top 20 of the local App Store sports Top Grossing chart, and treat that as an early signal that someone pays, not as a revenue forecast.
When to launch: football fixtures run all year, with no hard window. The 2026-27 season's major leagues are in full swing; launching before the winter break (mid-December) captures one more stretch of fixtures.
What would show the call was wrong: if VARStats has passed 50k Play installs one or two months from now, with a normal share of positive reviews and people in the comments mentioning "public accuracy," that would show the "transparency" differentiator really does pull users, and the grade should go up. Conversely, if after it stops buying ads its Uzbek and Russian chart rankings drop out of the top 50 within a week, the earlier payments were all bought with ads, and the playbook itself should be discounted.
Ads it's running
Google Seen 1 times, 1 still running, first seen Nov 2026 17.
These are the longest-running ads in this set. Long runs usually mean the creative is working. Data comes from the official Meta, Google and TikTok ad libraries and updates daily; open the official library to view the original creative.
Research notes: competitor table, acquisition evidence, full research report
- Competitors
Name Where Launched Scale How it differs from this one Forebet Web Play 2014 Play 3.58M installs, an established math-based prediction site with a Russian version Monetizes through ads and affiliate links, with no odds deviation or calibration record; VARStats runs no ads and charges by subscription FootyStats Web Play 2021 Play 708k installs; iOS Russia Sports Top Grossing #6 Stats-heavy database, mostly English; VARStats leads with probabilities vs. odds, goes deeper on Russian and Turkic AiXTIP iOS 2026-04 Play 17.7k installs; iOS Russia Sports Top Grossing #8 Also chasing Russian and Turkish users, with "AI prediction" pitches; VARStats sells verifiable calibration GG AI Football Predictions iOS 2026-06 Play 30k installs; iOS Uzbekistan Sports Top Grossing #7 "AI guaranteed win" style; VARStats adds bias to probabilities and promises no wins 11XScore AI Football Android 2025-08 Play 3.08M installs, 7-day +53k A heavily advertised big AI score-prediction app, mainly English and Portuguese markets 180Score AI iOS 2024-03 Play 2.94M installs; iOS Uzbekistan Sports Top Grossing #14 Same category, much larger, no public calibration BetMines Android 2019 7.16M Play installs A betting tips community that funnels to betting; VARStats has no affiliate links TipsPro AI, Scouter, GoScore AI and other Turkish "AI tahmin" apps Android 2021–2026 Anywhere from a few thousand to 1.1M installs, all in the top 200 of Turkey's Sports Top Free (TipsPro #30) Gives direct betting picks; VARStats gives full-market probabilities and odds bias Sofascore / Flashscore / FotMob Big-company features Already live Sofascore, Play 108M installs Only scores and stats, no probabilities - Where users come from
- Verified Meta (Instagram / Facebook) ads: Over a dozen ads under the "Varstats" page in the Meta Ad Library: 09-18 started with website ads; October Turkish creatives, some running only a few hours to test them; 11-05 and 11-06 brought Russian and Turkish "free download" app ads, and 11-17 added new ones. The timing lines up with the Play and iOS chart spikes.
- Verified Programmatic SEO across 16 languages (Yandex): The sitemap has about 36,784 URLs, each page carries 16 hreflang tags; robots.txt includes a Clean-param (yclid / ysclid) that only Yandex honors. But it's not in Tranco's top 1M, and CrUX places it only in the 2026-10 500k bucket, in Turkey.
- Verified Indie launch and directory sites: Fazier (41 votes), Microlaunch, Smol Launch, SaaSCity, Founder.best, SaaSRove, F6S and Crunchbase all have pages, mainly for backlinks.
- Verified Owned content (Sports.ru blog / Instagram / TikTok / Telegram bot): Two Sports.ru posts with +5 and +7 likes; a couple dozen likes on a single IG post; 1 TikTok followers; the @Varstatsbot bot. Effect negligible.
- Inferred Storefront category charts and store search: On iOS, the Sports category charts in Uzbekistan, Russia, Greece and Portugal have low entry barriers, so charting brings organic exposure. This is a guess with no direct evidence.
Why VARStats (com.varstats.app) is getting traction
Bottom line first
VARStats is a pre-match football probability analysis product built by Merdan Allaberdiyev, an indie developer in Ashgabat, Turkmenistan. It's available as a website, Android and iOS app, and Telegram bot, monetized by subscription. Calling it a breakout is premature: it has only 1,066 installs on Play worldwide, adding 70–80 a day, and has only reached #131–186 on Turkey's Play Sports Top Free. But its iOS version tells a more interesting story than Play: it hit #3–11 on Uzbekistan's Sports Top Grossing and peaked at #14 on Russia's Sports Top Grossing, which suggests real people are paying, though these storefronts have low barriers.
Broken down, its growth playbook has three moves and nothing viral:
- Small Meta (Instagram / Facebook) campaigns. Website ads started in Sep 18; in Nov 5 it switched to "free download app" ads, with creatives in Turkish, Russian and English. The Play install curve and the Turkish chart entries only appear after that.
- Website first, programmatic SEO in 16 languages. The sitemap contains roughly 37k URLs (match pages, team pages, league pages), each with hreflang tags for 16 languages. robots.txt includes a Clean-param that only Yandex recognizes, clearly aimed at Russian-language search.
- Targeting languages and payments that nobody serves well. Russian is the default language, with Turkish, Turkmen, Uzbek, Kazakh and Azerbaijani also supported. Russian users can pay via Mir cards or SBP on the web. Western English-first prediction apps don't bother with any of this.
The differentiation is mainly in positioning: it doesn't sell "guaranteed winners," it sells "probabilities plus a calibration record". It lines up model probabilities against implied odds to show the bias, publishes daily pick archives and calibration curves, runs no ads and has no betting affiliate links. That makes it look more credible among the many "AI guaranteed win" tips apps in Russian- and Turkic-speaking markets. But its current scale comes mainly from paid acquisition plus low competition in smaller languages, not from the differentiation itself having been proven by the market.
1. Basic data
| items | Numbers | Source |
|---|---|---|
| Domain registration | 2026-09-14 | Verisign RDAP |
| First Instagram post / first Meta ad | 2026-09-18 | Instagram @varstatscom, Meta Ad Library |
| Earliest model update log entry | 2026-09-25 | varstats.com/en/method |
| Listed on Android | 2026-10-21 | googleplay library app.released |
| iOS release | 2026-10-28(id6818466525) | iTunes lookup |
| Play cumulative installs | 412(11-11)→ 551(11-12)→ 618(11-14)→ 838(11-17)→ 1,066(11-19) | googleplay library app_snap |
| Play charts | Turkey SPORTS Free chart only: 186(11-11) → 131(11-15) → 164(11-19) | googleplay library chart_rank |
| Play rating | None | googleplay library |
| Play in-app purchase range | $2.99–$49.99 | googleplay library |
| iOS Uzbekistan | Sports Top Free 26–58 (since 11-07); Sports Top Grossing 4(11-13), 11, 6, 8, 3(11-17, 11-18), 9(11-19) | appstore library, chart_rank |
| iOS Russia | Sports Top Grossing 31(11-05) → 112(11-12) → 14(11-16) → 44(11-19) | appstore library |
| iOS other | Greece Sports Top Grossing 36–82 (since 11-15), Portugal Sports Top Grossing 83–186, Turkey Sports Top Grossing 88–172; also seen on the Sports Top Free charts of a dozen-odd smaller storefronts such as Bulgaria, Cyprus and Malta | appstore library |
| iOS ratings | 2 entries in Uzbekistan, 2 in Russia, 0 in the US store | appstore library / iTunes lookup |
| Website traffic | Not in Tranco's top 1M; CrUX places it only in the 2026-10 500k bucket, in Turkey | traffic library |
| Pricing | $3/week, $5/month, $38.40/year (list price $48, promo code VAR20 through 11-30); 3-day free trial, no card required; Mir / SBP available in Russia; also accepts one-time crypto payments | varstats.com/en/pricing |
How to read these numbers:
- On Play it gets about 70–80 installs a day, a very small scale. Reaching #164 on Turkey's Sports Top Free only takes about a thousand installs, which shows how low that chart's bar is. On the same chart on the same day, more than twenty Turkish-language "AI prediction" apps sit alongside it (TipsPro AI at #30, Scouter at #39, GoScore AI at #40, AiXTIP at #173…).
- The revenue signal is on iOS. It's #3 on Uzbekistan's Sports Top Grossing, behind names like Setanta Sports, Tennis TV, UFC and Flashscore; #14 in Russia, near FootyStats (#6) and AiXTIP (#8). But Sports Top Grossing in both storefronts has a very low bar (in Russia Apple only supports carrier billing), so a handful of subscriptions can push a app up the ranks. So "people are paying" is probably true, but "it's earning a lot" doesn't follow.
II. Competitors
Plenty of products do the same thing ("football match probabilities / predictions + subscription"), and this category is crowded:
| Product | Where | Launched | Scale | How it differs from VARStats |
|---|---|---|---|---|
| Forebet | Web + Android | 2014 | Play 3.58M installs; the website is an established math-based prediction site with a Russian version | An established "mathematical prediction" site, monetized through ads and affiliate links; no odds bias or calibration record |
| FootyStats | Web + iOS + Android | Play 2021 | Play 708k installs; iOS Russia Sports Top Grossing #6 | Stats-focused database that also offers predictions and Premium; mostly English, with Russian and Turkish support |
| BetMines | Android | 2019 | 7.16M Play installs | A betting tips community with affiliate and betting funnels |
| AiXTIP | iOS + Android | 2026-04 | Play 17.7k installs; iOS Russia Sports Top Grossing #8; Turkey Play Sports Top Free #173 | A Turkish developer with an "AI prediction" pitch, competing with VARStats for the same Russian- and Turkish-speaking users |
| GG AI Football Predictions | iOS + Android | 2026-06 | Play 30k installs; iOS Uzbekistan Sports Top Grossing #7 | "AI guaranteed win" messaging; in Uzbekistan's chart it ranks right next to VARStats |
| 11XScore AI Football | Android | 2025-08 | Play 3.08M installs, +7k over 53 days; Brazil Sports Top Grossing #10 | A heavily advertised big "AI score prediction" app |
| 180Score AI | iOS + Android | 2024-03 | Play 2.94M installs; iOS Uzbekistan Sports Top Grossing #14 | Same category, far larger scale |
| TipsPro AI(İddaa Tahminleri) | Android | 2024-09 | Play 329k installs; Turkey Sports Top Free #30 | Local Turkish İddaa predictions, positioned as "we tell you what to bet on" |
| Scouter, GoScore AI, Skor Analiz AI, ForeKick AI, and others | Android | 2021–2026 | Ranging from a few thousand to 1.1M installs, all in the top 200 of Turkey's Sports Top Free | A wave of Turkish "AI tahmin" apps appeared in 2026 |
| ScoreGPT, Sport AInsights, TejTips, HockeyPredicts, and others | iOS + Android | 2026 | Around #23–45 on Russia's Sports Top Grossing | Subscription prediction apps for Russian-speaking markets |
| Statz.ai, Tiki Taka, fottips, and others | Web | 2025–2026 | Not verified | English-market "AI match preview" site |
| Sofascore / Flashscore / FotMob | Big-company app | Already live | Sofascore, Play 108M installs | Scores and stats, no win probabilities. It's an app people already have on their phones. |
Verdict: there's no empty market, and the competition is crowded. VARStats is a late entrant, not a pioneer.
3. Where users come from (growth levers)
1. Meta ads (confirmed, the main channel)
Searching "varstats" in the Meta Ad Library shows a dozen-plus ads under the "Varstats" page (captured 2026-11-19):
- 2026-09-18 → 10-13: the earliest one, pushing the website.
- Between 10-09 and 10-15 there are a few that only ran for 1–3 days (one shows "Total active time 3 hrs"), which looks like creative testing.
- 10-13 → 11-05 and 10-15 → 10-29: Turkish creatives, such as "Maç başlamadan önce veriye bak… 3 gün tam erişim ücretsiz. 👉 varstats.com" and "Sezgiler yanılabilir. Rakamlar konuşur… 18+ · Olasılıklar garanti değildir."
- 11-01 → 11-11: English "Football probabilities for 100+ leagues… Live scores on your lock screen. Our accuracy is public."
- At least 4 ads running since 11-05, 11-06, including Russian "…115 лиг и открытый учёт точности. Скачайте VARStats бесплатно." and Turkish "…VARStats'ı ücretsiz indir.", both app download ads.
- 11-17 brought a new one, still running.
The timeline lines up: App install ads started 11-05, the Russian App Store sports Top Grossing chart saw it at #31 on 11-05, the Uzbek sports Top Free chart pushed it up to #27 on 11-07, and it showed up on the Google Play Turkey chart on 11-11. Our own metaads library doesn't track it, so we can't see how much was spent. Source: facebook.com/ads/library/?q=varstats
2. Programmatic SEO on the website (confirmed as being done, results not verified)
- varstats.com's sitemap has 13 shards, about 36,784 URLs, and each page carries hreflang for 16 languages: ar / az / de / en / es / fr / hi / it / kk / pt / ru / tk / tr / uz / vi / zh. Page titles follow one format: "X vs Y: prediction and stats", "Albacete - Sabadell: bepul prognoz", "Braga - Sporting Lizbon: tahmin ve istatistik".
- robots.txt contains
Clean-param: … yclid&ysclid …(only Yandex honors this), which suggests a deliberate push for Russian-language Yandex SEO. It also blocks AhrefsBot and allows PerplexityBot. - Brave returns plenty of competition pages (site:varstats.com). But the site isn't in Tranco's top 1M, and CrUX only places it in the 2026-10 bucket in Turkey (500k bucket), so search traffic is still small.
3. Indie product launch sites and directories (confirmed, the traffic they bring is negligible)
Listed on Fazier (41 upvotes), Microlaunch, Smol Launch, SaaSCity, Founder.best, SaaSRove, F6S, Crunchbase, and PromoteProject. The Fazier page has the founder's own intro post, and the comments are mostly mutual upvoting. These mainly earn backlinks and SEO authority, not gambling users.
4. Owned content (confirmed, almost no effect)
- On the Sports.ru blog "VARStats футбольная статистика": 11-02 posted "checked 281,621 predictions, the model said 70% and it happened 72.6%", and 11-09 posted "Spain vs England, the model gives the home side 59%". They got only +5 and +7 likes.
- Instagram @varstatscom: reels since Sep 18, tagged "AI content", with about two dozen likes per post.
- TikTok @varstats_official: 1 followers, video views close to 0.
- 0 mentions on X (x-radar shows no results in the last 60 days). Nothing on Reddit.
- Telegram bot @Varstatsbot: "Вероятности матчей, «Выбор дня» и ваши команды".
5. Store search / small-storefront charts (inferred)
On iOS in Uzbekistan, Russia, Greece, and Portugal, the Sports category's Top Free and Top Grossing charts have low thresholds, so just charting brings some organic exposure. iOS supports 16 languages, and the store page is localized to match.
4. Why it took off
- Paid ads bought the first batch of users directly. Meta ads pushed the website from mid-September, then switched to app download ads in Nov 5. The Play and iOS chart signals show up after that.
- It picked small-language markets with thin competition and real willingness to pay. Russian is the default, plus Turkmen, Uzbek, Kazakh, Azerbaijani, and Turkish, which are the founder's own language circle. Football betting is hot in these markets, but few Western prediction apps localize seriously. That's how it got into the top 3 of the Uzbekistan iOS Sports Top Grossing chart and #14 in the Russia storefront.
- It kept a payment path open for Russian users. Since 2022, Google Play no longer takes payments in Russia, and Apple only offers carrier billing. VARStats added Mir card and SBP payments on the web, plus cryptocurrency, catching money other apps can't.
- The website came first, with multilingual SEO, laying groundwork for long-term organic traffic (no clear traffic spike yet).
- Web first, one account across all platforms: a subscription bought on the website also works in the app, which avoids part of the store commission.
5. Differentiation
- Probabilities plus odds deviation: each market shows "model probability / bookmaker implied probability (margin removed) / Δ" side by side, helping users spot where the model and the market disagree.
- Public calibration and archive: the "Pick of the day" is written into the archive before kickoff and never edited. In the last 30 days, 346 of 246 matches hit (71%), and overall 653 of 470. Calibration results for 280k predictions were posted on Sports.ru. The model changelog runs from Sep 25 to Nov 11, with 11 entries.
- No ads, no affiliate links: most competitors make money through betting affiliate links, and it explicitly doesn't.
- Fine-grained market coverage: corners, yellow and red cards (adjusted for referee strictness), shots, fouls, half-time, Asian handicap, and season simulations, across 114 competitions.
- Localized language and payments (see above).
Is this why it took off? Partly. What actually brought users in was paid ads and small-language localization. "Transparency plus calibration" gives it something to say in the ad creatives ("Our accuracy is public", "открытый учёт точности"), and may raise the trial-to-paid rate, but there are no reviews and no word-of-mouth yet, so the evidence isn't strong enough to say users came for it. On iOS in Russia and Uzbekistan, its direct rivals AiXTIP and GG AI use "AI sure wins" pitches and also reach the Top Grossing chart.
VI. Visibility
Few people know about it yet: revenue isn't public, there's no TrustMRR listing, and no founder has shared revenue. No media coverage. 0 mentions on X. 1 TikTok followers. The two Sports.ru blog posts got +5 and +7 likes. The only traces are a pile of indie launch-site pages and Meta ads.
Copycats: none seen. It's itself a late arrival in the "AI football prediction" trend.
VII. What others can borrow
- Language selection: pick language markets the developer knows and Western products don't bother with (Russian plus Central Asian Turkic languages), then add payment methods that work locally (Mir / SBP, crypto).
- Website first, programmatic SEO: one page per match, hreflang across 16 languages, with Clean-param tuned for Yandex. The app is just another entry point to the website subscription.
- Make "verifiable" the selling point: public archive, calibration curves, and model changelog stand out in a category full of "guaranteed win" pitches.
- Small-budget Meta ad rhythm: test creatives on the website first (some ran only a few hours), then switch to app install ads, with a separate creative set per language.
- Top Grossing charts in low-threshold storefronts can serve as an early paid-demand signal, but don't use them to infer revenue scale.
8. What's still uncertain
- Actual revenue and paid conversion rate: the Uzbekistan and Russia Sports Top Grossing charts have low thresholds, so a ranking only shows that someone is paying.
- How much was spent on Meta, which countries it targeted, how many installs it drove, and whether Uzbekistan iOS downloads came from ads or store search.
- How much organic search traffic the website actually gets (Tranco and CrUX both show very little).
- Whether Play installs drop off once ads stop. No ratings on Play, and retention is unknown.
- A 30-day 71% hit rate on recommendations that require odds of at least 1.30 doesn't guarantee profit (the published metric is hit rate, not return on investment).
Sources
- Data card: /home/ljr/fanout/why/cards/play_com.varstats.app.json
- googleplay library (app, app_snap, chart_rank, app_profile, data through 2026-11-19)
- appstore library (chart_rank, app_profile, v_chart, data through 2026-11-19)
- traffic database (rank_day, crux_domain, product_domain)
- metaads library (this advertiser isn't tracked)
- varstats.com/en, /en/pricing, /en/method, /en/stavka, /robots.txt, /sitemap.xml (read 2026-11-19)
- www.facebook.com/ads/library/?q=varstats (read 2026-11-19)
- fazier.com/launches/varstats
- www.sports.ru/tribuna/blogs/varstats/, www.sports.ru/football/blogs/3594712.html, www.sports.ru/football/blogs/3618765.html
- www.instagram.com/p/DeJUfWqAO5Z/, www.tiktok.com/@varstats_official, t.me/Varstatsbot
- varstats.com/en/privacy (developer address: Ashgabat, Turkmenistan)
- itunes.apple.com/lookup?id=6818466525
- Verisign RDAP (varstats.com registered 2026-09-14)
- x-radar search for "varstats" over the last 60 days: 0 results
