(function(){try{ var IS_CINE=false; var AST_FIRST=4,POP_GAP_MS=17000,AAB_FLAG="__cineAabDone",AAB_EVENT="cine:aab-done"; var LK="cine:popAt",HOP="/api/prefs/frame",AK="cine:aabDay",AF="cine:aabFirst"; function skip(){var p=(typeof location!=="undefined"&&location.pathname)||"";return /^\/admin(?:\/|$)/.test(p)||/^\/[^\/]+\/(?:premium|layout-probe)(?:\/|$)/.test(p)} /* A scraper session, not a visitor. The app only ever writes ?season=N&episode=N (TitleClient's two qs.set calls, in that order); the residential-proxy catalog scrape lands episode-first on 80-98% of its requests, one request per IP, in a real browser that runs this script. On 2026-09-09/10 it put 35-45k such "users" a day into Adcash's zone 12091114: 7.4k users and $12.36 on the 7th became 57.8k users and $2.27, with the unflooded child zones losing 40% too. These sessions must never reach a network's tag or the visitor beacon. Real episode-first landings are ~0.4% of visitors and lose ads on that one pageview, which is the whole cost. */ function scraped(){var q=(typeof location!=="undefined"&&location.search)||"";return /[?&]episode=\d+&season=/.test(q)} function adsOff(){try{return /(?:^|;\s*)ads_off=1(?:;|$)/.test(document.cookie||"")}catch(e){return false}} function ls(k){try{return localStorage.getItem(k)}catch(e){return null}} function ss(k,v){try{localStorage.setItem(k,v)}catch(e){}} function last(){return Number(ls(LK)||0)||0} function mark(){ss(LK,String(Date.now()))} function today(){return new Date().toISOString().slice(0,10)} /* No tags on this page: the client-side zone (if mounted later) has nothing to wait for. */ if(skip()||scraped()||adsOff()){aabDone();return} /* Adsterra first, then the Adcash zones, one after another (2026-10-27, the user: the pops overlapped). Adsterra's tag lays its own full-page layer (a div with a random id, z-index 2147483647, opacity 0.01) and every armed Adcash zone lays one too (#dontfoid), so with two up a click fed one network and the other's pop was lost. Measured on a residential exit: each press on Adsterra's layer is one pop, its own first-party counter pp_sub_ goes 1,2,3,4, pp_delay_ holds a ~10 s gap, and after the 4th there is no layer until its 2 h period ends. Adsterra is done for this page once its tag failed to load (403 on datacenter addresses, an adblocker), it has popped AST_FIRST times this period, or a press went by with no pop while it was not in its gap (no ad for this visitor). Only then does the anti-adblock zone's runPop run (its lib loads at once), and only once that zone has popped or had no ad does the boot hand on to the client-side zone (AAB_FLAG/AAB_EVENT, components/layout-mesh/AdcashPop). Cookies and the DOM are read; nothing of theirs is touched. */ var AST_SETTLE_MS=2000,AST_PRESS_MS=1500,AAB_ARM_MS=5000; var astIsDone=false,astWait=[]; function astDone(){if(astIsDone)return;astIsDone=true;var w=astWait;astWait=[];for(var i=0;i0)){astDone();return} var loadedAt=0; function pops(){return Number(ck("pp_sub_"+key))||0} function gap(){return ck("pp_delay_"+key)!==null||ck("pp_idelay_"+key)!==null} function capped(){if(pops()>=AST_FIRST)astDone()} s.onload=function(){loadedAt=Date.now()}; s.onerror=astDone; capped(); if(astIsDone)return; /* A press in the first moments after the tag loaded may land before its layer is up: it says nothing. */ try{document.addEventListener("click",function(e){ if(astIsDone||!e||!e.isTrusted||!loadedAt||Date.now()-loadedAt=AAB_ARM_MS)){aabDone();return} setTimeout(tick,250)})(); } /* At least POP_GAP_MS between two Adcash pops (2026-10-27, the user: "ensure there's really 17s between ads"). A press while an Adcash layer is up ([znid]) is a pop -- its lib pops on any press while armed, measured; its time goes to localStorage, so every tab and page load sees it. Inside a page Adcash re-arms on its own timer (18.3-19.5 s after a pop, one zone, residential exit); what could come sooner is a reload or a new tab calling runPop afresh, so runPop waits until the gap is up. Their listeners sit on window in the capture phase, ahead of the page's own, so a press cannot be held from them without holding the site's clicks too: this times our call and reads, nothing more. */ var POP_AT="cine:adcPopAt"; function popPress(e){try{if(e&&e.isTrusted&&document.querySelector("[znid]"))ss(POP_AT,String(Date.now()))}catch(_){}} try{window.addEventListener("pointerdown",popPress,true)}catch(e){} function popAfterGap(fn){var w=(Number(ls(POP_AT))||0)+POP_GAP_MS-Date.now();if(POP_GAP_MS>0&&w>0&&w<=POP_GAP_MS)setTimeout(fn,w);else fn()} /* Adcash anti-adblock: plain tag after a delay, then one runPop once Adsterra is done (above). Frequency is Adcash's own dashboard cap, per zone — the zone IS the rung of the frequency ladder, so we never gate it ourselves. The 90s delay is identical on every rung so it cannot become a second variable. */ function runAab(aabPath,zone,aabDelay){ if(!aabPath||!zone||document.getElementById("cine-adcash-aab")){aabDone();return} setTimeout(function(){ try{ if(adsOff()){aabDone();return} var a=document.createElement("script"); a.id="cine-adcash-aab"; a.src=aabPath; a.async=true; a.onerror=aabDone; a.onload=function(){afterAst(function(){var t=0;(function boot(){var lib=null;try{lib=window.aclib}catch(e){} if(lib&&typeof lib.runPop==="function"){popAfterGap(function(){try{lib.runPop({zoneId:zone})}catch(e){aabDone();return}aabWatch(zone)});return} if(t++<120)setTimeout(boot,50);else aabDone()})()})}; (document.head||document.documentElement).appendChild(a); }catch(e){aabDone()} },aabDelay); } /* AdMaven anti-adblock. Their PHP echoes the payload inline at the top of with no delay, so this injects it as early as the boot script runs and never gates it: frequency is AdMaven's own dashboard cap, exactly as with Adcash. No delay knob on purpose - a 90s one is what cost Adcash half its capture, and there is no reason to repeat the experiment. */ function runAdmvn(p){ if(!p||document.getElementById("cine-admaven-aab"))return; try{ var s=document.createElement("script"); s.id="cine-admaven-aab"; s.src=p; s.async=true; (document.head||document.documentElement).appendChild(s); }catch(e){} } function aabBeacon(arm,ev){try{var b=JSON.stringify({arm:arm,event:ev});if(navigator.sendBeacon){navigator.sendBeacon("/api/prefs/panel/event",new Blob([b],{type:"application/json"}));return}fetch("/api/prefs/panel/event",{method:"POST",body:b,keepalive:true}).catch(function(){})}catch(e){}} /* The arm may name a different Adsterra placement (astb runs Popunder_2, capped higher in their dashboard). An absent or non-Adsterra src falls back to this site's default placement, so a response from an older build - or a field the server chose not to send - still serves an ad. */ function astTag(src){ var u=(typeof src==="string"&&src.indexOf("https://")===0)?src:""; if(!u||document.getElementById("cine-adsterra-pop")){astDone();return} var key=astKey(u); astUncap(key); var s=document.createElement("script"); s.id="cine-adsterra-pop"; s.src=u; s.async=true; astGate(s,key); (document.head||document.documentElement).appendChild(s); astWatch(key); } /* Adsterra pop counter (2026-10-10). The tag counts its own pops in a first-party cookie, pp_sub_ ( = the tag file's hash), renewed for its 2 h period on every pop; koddo's placement allows 4 per 2 h, 10 s apart. We only read it - see the header for why nothing of theirs is ever wrapped - and report each increase once: the last value seen is kept in localStorage, which every tab shares. Nothing is gated on the count; it exists to be held against the impressions Adsterra reports. */ var AST_SUB="cine:astSub:",AST_OLD="cine:astPops"; function ck(n){try{var a=String(document.cookie||"").split(";");for(var i=0;i=5){try{document.cookie="pp_delay_"+key+"=; max-age=0; path=/";document.cookie="pp_main_"+key+"=; max-age=0; path=/"}catch(e){}}} function astWatch(key){ if(!key)return; function check(){ var cur=Number(ck("pp_sub_"+key))||0,last=Number(ls(AST_SUB+key))||0; var k=cur>last?cur-last:(cur0)astBeacon("ast_pop",k)} try{document.addEventListener("click",function(){setTimeout(check,400);setTimeout(check,2000)},true)}catch(e){} try{document.addEventListener("visibilitychange",check)}catch(e){} try{window.addEventListener("pageshow",check)}catch(e){} try{if(typeof setInterval==="function")setInterval(check,15000)}catch(e){} check(); } /* Monetag's own snippet reads the zone off the script element's dataset before loading, so the attribute has to be set on the element it appends - not passed as a query param. Frequency is theirs to control, exactly like Adsterra's tag: we gate neither, or the comparison is unfair. */ function mtgTag(zone){ if(!zone||document.getElementById("cine-monetag-pop"))return; var s=document.createElement("script"); s.id="cine-monetag-pop"; s.dataset.zone=String(zone); s.src="https://llvpn.com/tag.min.js"; (document.head||document.documentElement).appendChild(s); } /* Sites outside the experiment run a fixed shape: Adsterra's own placement when the site has one, Adcash, plus AdMaven once its tag lands. No arm and no beacons - mixing their traffic into aab_arm_daily would corrupt the denominator of the running ladder, and a split test needs a control to split. */ if(!IS_CINE){astTag("");runAdmvn("");runAab("/0e2799e1cc907d4b.js","12369438",0);return} /* Adcash frequency ladder. One request decides the whole ad shape for this visitor: which Adcash zone (the rung) and whether Adsterra renders at all (control only). Both come from the server so a visitor can never be counted in one arm while being served another's zone. */ fetch("/api/prefs/panel",{credentials:"same-origin",cache:"no-store"}).then(function(r){return r.json()}).then(function(a){ if(!a){astDone();aabDone();return} var arm=a.arm||"control"; /* One visitor beacon per browser per UTC day: the denominator for revenue per visitor. */ var d=today(); if(ls(AK+":seen")!==d){ ss(AK+":seen",d); aabBeacon(arm,"visitor"); var first=ls(AF); if(!first)ss(AF,d);else if(first!==d)aabBeacon(arm,"returning"); } if(a.tag)astTag(a.tagSrc);else astDone(); if(a.monetagZone)mtgTag(a.monetagZone); if(a.admaven)runAdmvn("/"+String(a.admaven).replace(/^\/+/,"")); /* Number(0)||90000 is 90000, so a deliberate zero delay has to survive the coalesce explicitly. */ var dly=Number(a.delayMs);runAab(a.file?"/"+String(a.file).replace(/^\/+/,""):"",a.zone,Number.isFinite(dly)&&dly>=0?dly:90000); }).catch(function(){astDone();aabDone()}); /* The direct-link broker (Adcash / Monetag / AdMaven) runs identically in both arms. */ fetch("/api/prefs/hint",{credentials:"same-origin",cache:"no-store"}).then(function(r){return r.json()}).then(function(d){ if(!d||!d.ok)return; var cool=Number(d.cooldownMs)||600000,wait=Number(d.firstDelayMs)||20000,t0=Date.now(),fired=0; function ready(){return Date.now()-t0>=wait&&Date.now()-last()>=cool} function fire(e){ if(fired)return; if(!e||!("isTrusted"in e)||!e.isTrusted)return; if(adsOff()||!ready())return; fired=1; mark(); try{window.open(HOP)}catch(err){} setTimeout(function(){fired=0},cool); } document.addEventListener("pointerup",fire,true); document.addEventListener("click",fire,true); }).catch(function(){})}catch(e){try{window["__cineAabDone"]=1;window.dispatchEvent(new Event("cine:aab-done"))}catch(_){}}})();

The Fisher

Comments

You may also like

6 Below: Miracle on the Mountain

Stillwater

The 355