The Expected Kill Philosophy.
Using data collected from hundreds of killbets and machine learning models to predict whether Goorilaa will win his killbet.
Warzone is returning and so are we!
After a successful initial project in the Battle Royale/WZ scene, we have decided to extended it into WZ2023/24.
Data-driven decisions. Seize opportunities.
Partnered with
@GoorilaaCapital
to apply our predictions directly into a fund, maximising earnings.
Currently have 5 or so people invested into the scheme.
DM for Discord watch parties, in-depth statistical analysis or just questions regarding the fund.
Hit a setback today which we couldn't explain (back-to-back incorrects).
Before I operated on a 3-hour period, I have updated the coefficients to an hourly basis.
I have also added coefficients for weekends and weekdays to reflect lobbies during those periods respectively.
Goorilaa has moved to doing duos with Calhouns.
Killbet: 25/2 w/ G Willie
Outcome: Goorilaa to win.
There is very little data on duos, and so a lot is inflated.
xK ~ 26, we're expecting it to be close.
Insights from WZ3: Day 1
WZ3 allows for higher clutching opportunities during map 2 due to SBMM and low-kill nature of the game.
Goorilaa KBs involving Huel and Willie suggest 25/2 duos and 38/2 trios are balanced numbers however are due to increase with gametime.
"People are dumb for coming to him, he makes no one any money, I could guess."
Before I monetized the data and insights, ask our top fan how good we were 😎
Despite regularly underachieving map 1, map 2 often is more successful for obtaining high kills.
Clutch factor has been reduced from 47% to 44% of required kills from WZ2.
Goorilaa managed to get the highest number on the wheel (again!).
We were transparent with Goorilaa and Will and explained 42/2 would be the best balance between chat and themselves for a competitive, fair killbet. They achieved 43, just 1 more than we would predict.
Killbet: 36-45 w/ G Huel Willie
Outcome: Goorilaa to win.
The potential xK for this squad is 42 and on average they are more likely to reach their potential during this time (04:30 AM to 05:00 AM GMT +0:00).
In trios, they have a win rate of 50% at required kills of 41.89.
What factors affect our predictions to make them unreliable?
1. Jailbreaks - These occur more often than not when you have $ on.
2. Undetectable Hackers - They skew data.
3. Resets - These don't affect the model too badly if it happens during map 1.
Goorilaa (+ team mates) are more likely to reach their potential when playing for higher amounts.
Their potential is defined as the maximum amount of kills possible, and is calculated by averaging their best games. The % difference is how close they are to reaching it.
First KB of the day... and we believe Goorilaa is being audacious.
Killbet: 50-55 wheel w/ G Huel Willie
Outcome: Goorilaa to lose.
None of our xK values exceed 50, except for the highest potential value (53.9). Time factor reduces this to ~ 49.
Following on from our profit-loss analysis, we can confirm chat are in a LOSS of nearly $14,000.
This does NOT necessarily mean that Goorilaa or anyone in particular is "up" on chat however it does show chat make poor decisions more often than not.
1/2
Slow and steady wins the race and the same applies to kill bets, here at Goorilaa Capital Partners we believe in responsible informed betting. 🧠
We encourage you to wait for the right time to bet, there will always be more opportunities! 🧐
1/2
What a prediction! This game had a jailbreak, Mike Tyson and a KO to Goorilaa's chances at turning a profit today.
We're so glad we went with our model, and now trust it to make decisions when we can't justify it with loads of data like we usually do.
3/3 ✅✅✅
🔥 New KillBet Alert 🔥
BET: 85/2 w/ G HUEL R1 (NA Host Only, Trios)
Outcome: Goorilaa to lose.
We honestly have no idea with this one, we can't justify the 3/3 "lose" signals by our models. I guess it's time to see if we can trust them.
Bet low & cautiously, we aren't.
New Trackable Metrics:
Individual Player KPIs - Kills, Deaths & K/D (overall for duos/trios/quads)
Game KPIs - Time of Day, Initial Dropzone
Social KPIs - Sensitivity Analysis, Viewership
First KB of the day... and we believe Goorilaa is being audacious.
Killbet: 50-55 wheel w/ G Huel Willie
Outcome: Goorilaa to lose.
None of our xK values exceed 50, except for the highest potential value (53.9). Time factor reduces this to ~ 49.
We've improved our 'Expected Kills' calculation by factoring in Jailbreaks & early resets.
This can be used by anyone to determine the favorability of a bet in upcoming streams by simply adding up the "xK" for each player involved.
Anyone can use it.
Should we release? 👀
🔥🔥New KillBet Alert 🔥🔥
BET: 105/2 w/ G HUEL WILLIE R1
Outcome: Goorilaa to lose.
RFC: "No"
DT: "Yes"
SVM: "No"
2/3 models predict Goorilaa to lose, and this is a step up from the previous 100/2 completed by the squad, plus 105/2 was failed by better squads.
@Goorilaa
@ClutchRayn
@TwizzSK
We'd like to add we are a neutral party that collects statistics on killbets, and are able to detect whether foul play has occured with very high accuracy.
Nothing was detected during this killbet. The accusations are false.
Killbet: 50/2 w/ G Huel Willie
Outcome: Goorilaa to win.
We assume this killbet will go beyond 3 AM, and so the xK significantly increases.
The xK ~ 53 across both maps assuming map 1 is played before 3 AM, and map 2 is after 3 AM.
We've just got into the stream.
Disrupt and Willie have a 100% win record together.
75-80 are quite low numbers, and did favour Goorilaa.
To even stand a chance, chat require an 80-85 wheel from this trio.
His early evening streams (18:00 to 22:00 GMT) also only have a win rate of 50% (3/6).
However due to little data available and the majority of data points being collected at the start of WZ3 in this range, this should be taken with a grain of salt.
People think KBs are only about kills. They're not.
Our insights have derived the sweet spot for betting on Goorilaa lies between 22:00 and 02:00 (GMT +0:00), where he is most likely to lose the KB.
This time is usually accompanied during the first two-three KBs of the stream.
For example,
For split hosts, Goorilaa has a win rate of ~ 68.9% across duos, trios & quads.
After 6 AM, bets are hard to place due to extremely higher than average win %. However, between 10PM - 2 AM there may be potential depending on xK, as win % is ~ 4-13% lower.
Chat are down horrendous, the most they've been down since the start of WZ3.
If you followed all of our tips and advice you'd be up money, not down like the rest of chat.
Follow data-driven insights and models to maximise your earnings.
🔥 New KB Alert 🔥
BET: 100/2 w/ G HUEL WILLIE WIZZ (split host/quads)
Outcome: Goorilaa to win.
Low kills for this squad, considering they smashed 100/2 with R1, 107/2 with Disrupt, 98/2 with Disrupt.
Models suggest they achieve this too, with 2/3 saying "Win" (for G).
Furthermore, I do not believe Goorilaa or anyone he plays with gives into over 40% of killbets just for "content".
His win rate is currently at 58.76%. Ours is over 80%, most of which does not result in a loss of money. Who is the better guesser really?
Another KB involving Shifty? You already know the outcome guys...
BET: 95/2 w/ Goorilaa Gorgo Shifty
Outcome: Goorilaa to win.
They already smashed 96/2 at an earlier time, and no number seems to be a problem with Shifty. He completed 110/2 with Lenun and G recently.
3/3 on the day.
Prediction was correct yet again and as more data piles in, we can convey predictions with more confidence.
Chat is still down on the day, but not by a lot.
🔥 New Prediction Alert 🔥
BET: 130/2 w/ G DISRUPT DUHVERA ABWIZZ (split host)
Outcome: Goorilaa to lose.
Here's what we know:
3/3 models return 'Loss' for Goorilaa.
This team has an expected kills of exactly 130.
We trust our models, do you?
We've just seen Hueligan get banned on stream for potential violations, and he is under review.
We have also noticed significant performance increases statistically, however haven't illustrated them as our focus has been on making correct predictions.
Final prediction before we head offline.
Killbet: 50/2 w/ G Huel Willie
Outcome: Goorilaa to win.
xK dropped to ~ 54, however if pushed past 6 AM the chance of chat winning drops significantly.
xK more reliable since split hosts are being used, rather than NA only.
Another day another killbet.
BET: 80/2 w/ G Willie Disrupt Trios (EU only).
Outcome: Goorilaa to lose.
Only 1 dataset in the data book, if no resets happen prediction should stand.
RFC: 'No'
SVM: 'No'
DT: 'No'
3/3 models predict no for trios. Stay cautious as 79/2 was a W.
We're guiding our investors well through "No Flare" rules in killbets.
Just now we predicted an xK of 51 and they got 50, allowing the fund to make a strong return after a long dry period of no viable bets or numbers.
Another clinical victory for us.
Thanks
@GoorilaaCapital
!
We have once again protected our investors from rigs by Goorilaa 🛡️
Chat is now down ~ 27,500£ net (detailed statistics to come at the end of the stream).
Story of the first 21 killbets (3 streams worth).
The average WZ player is up £482.09.
The chat as a whole is down £979.00.
The chat has only hit positive balance 4/21 killbets.
We're working on bringing accurate profit-loss data from Goorilaa's killbets, even though they have no impact on our models.
We want to know who's really winning.
Here are some top expected earners:
*Assuming pots are evenly divided after wins/losses*
Killbet: 45-50 w/ G Huel Willie
Outcome: Goorilaa to lose.
The xK is 46, chat favoured 60/40.
Not as clinical as previous bet, however the information is there should you wish to bet.
Hedge fund has sat this one out, we only look for clinical wins.
We've reached 100 KBs.
DMs are open for any insights/statistics for any particular player.
We're 15/63 predictions incorrect.
10/15 resulted in NO monetary loss (overestimating capabilities).
2/15 detected "foul play" influencing results.
Looking to improve.
Killbet: 50/2 w/ G Huel Willie
Outcome: Goorilaa to win.
We assume this killbet will go beyond 3 AM, and so the xK significantly increases.
The xK ~ 53 across both maps assuming map 1 is played before 3 AM, and map 2 is after 3 AM.
Highest pot of WZ3 so far at over £5,000.
Killbet: 77/2 w/ G Skilli Disrupt Willie
Outcome: Goorilaa to win.
Purely analytical due to low data on this 4-man. We only have 15 data points for WZ3 so building prediction versatility within our machines is priority.
Killbet: 53/2 w/ G Huel Willie
Outcome: Goorilaa to win.
xK shifted to ~57 post 3 AM, however unsure as to how many times they've done double NA.
The risk of losing outweighs the statistical potential this trio has at this time.
Another winning & profitable pick for investors
@GoorilaaCapital
and wider independent bettors.
We're 4/4 since the beginning of the partnership in determining when Goorilaa.
We're up loads! DM for more information.
It's the xK influence.
No more rigs for Goorilaa, we're putting our foot down with numbers.
Chat has seen green only a handful of times, don't let today distract you from the bigger picture.
Working on adjusting to WZ3, there are several key differences:
Initial Lobby is 33.3% less than WZ2 which models haven't reflected thus far. Predictions remain extremely conservative.
Deposable buy-stores, flares & gulag tokens all reduce risk of players dying when reviving.
We've detected two statistical anomolies to suggest a VPN was being used this KB.
Across ~ 80 KBs (~ 160 maps), when an EU host has been employed (53 times) Goorilaa has pinged under 20 ms at all hours of the day.
Today he pinged 46 ms on EU (red).
Another brilliant process
@GoorilaaCapital
ensures we capitalized on a significant margin, despite having one of their best games.
xK ~ 54.
Maximum Potential was ~ 58.
Killbet Requirement ~ 65.
They Achieved ~ 58.
Good start to the weekend 💸
SVM model returns 'No' however more accurate RFC and DT models return 'Yes'. These models haven't had a lot of data to be trained on but upper limits of this squad lie at around 108-110, but probability of wheel landing on this is less than 33% - which is the odds Big G gave.
Killbet: 30/2 duo trios w/ Hueligang & Calhouns
Outcome: Huel to lose.
xK ~ 29, however all data points are trained on split EU/NA hosts prior to this. Unsure how double NA would affect them at this time. Bet low & with caution.
🔥 Prediction Alert 🔥
BET: 116-122 WHEEL w/ G DISRUPT DUHVERA ABWIZZ
Outcome: Goorilaa to win.
No data on this specific 4-man, so no models.
They have a strong potential of winning, with squads worse than this one getting over 50% of the required kills map 1 for KBs 110+/2.
Killbet: 36-45 w/ G Huel Willie
Outcome: Goorilaa to win.
The potential xK for this squad is 42 and on average they are more likely to reach their potential during this time (04:30 AM to 05:00 AM GMT +0:00).
In trios, they have a win rate of 50% at required kills of 41.89.
🔥❗️🔥 HIGH KB ALERT 🔥❗️🔥
130 in 2 w/ GOORILAA LENUN JUKEYZ AND SHIFTY
Models predict that Goorilaa will win.
This is the best squad Goorilaa has played with in 2023, but lobby limitation factors could cause issues.
Tune in!
Killbet: 53/2 w/ G Huel Willie
Outcome: Goorilaa to win.
xK shifted to ~57 post 3 AM, however unsure as to how many times they've done double NA.
The risk of losing outweighs the statistical potential this trio has at this time.
Just to confirm, we predict before map 1 starts and have never deleted a tweet, and always stand by them.
Go back through the Twitter and check.
But you guys know this already.
Across 28 killbet predictions, we've got 5 wrong, two of which detected foul play (for & against BigG).
That's a 82.1% success rate.
2 of these were from today, consecutively. We're looking into potential reasons as to why we overestimated their capability.
However, as most of the win % at these hours are much higher than 50% by ~ >10%, a strong case needs to be made for the xK vs the required kills in the killbet to place a bet.
Goorilaa has moved to doing duos with Calhouns.
Killbet: 25/2 w/ G Willie
Outcome: Goorilaa to win.
There is very little data on duos, and so a lot is inflated.
xK ~ 26, we're expecting it to be close.
Sorry chat, got this one wrong again.
4/6 on the day, major setbacks late into the stream.
Going to generate a method of aggregating Friday/Saturday/Sunday results separate to weekday results.
Final prediction before we head offline.
Killbet: 50/2 w/ G Huel Willie
Outcome: Goorilaa to win.
xK dropped to ~ 54, however if pushed past 6 AM the chance of chat winning drops significantly.
xK more reliable since split hosts are being used, rather than NA only.
"RIP killbets 28/12/2023"
We wonder why...
Chat hasn't statistically had a fair shot 20 from 25 most recent killbets & hit a new low of -£30,000 yesterday.
Killbet: 55/2 w/ G Disrupt Willie
Outcome: Goorilaa to win.
xK at this time for this trio is 61.
Goorilaa wins two thirds of his KBs between 3:00 AM - 6:00 AM GMT.
We've checked our data on this potential KB.
Goorilaa Skilli Huel Willie (0/2 in KBs).
xK as a team: 87 kills.
xK (summed individually for worst scenario, e.g. sweat lobbies): 94-95 kills.
Models on 85/2: 1/3 predict a 'Win' for Goorilaa with 70%+ confidence.
How accurate are our predictions?
Since WZ3:
Out of 16 public predictions, we have got 13 right.
Out of 1 private prediction, we have got 1 right.
Overall we're 14/17.
Spread the word, we're locked in.
Despite not making predictions for the remainder of the stream to avoid chalking pots for Goorilaa's sake, we still updated our data records.
Chat held a level. You guys don't need us 🤣
Blue = Training
Green = Correct Prediction
Red = Incorrect Prediction
White = No Prediction