50
◼SPORTS January 27, 2020more complicated than you couldeverevendreamorimag-
ine,” he says, adding that he’s notatlibertytosaymuchmore.S
chwimer, 33, pitched in 47 gamesin 2011 and 2012
before a shoulder injuryendedhiscareer.“Allmy
pitches were well belowaverage,”hesays.Tocom-
pensate,hemodeledwhichonestothrowandwhen,spend-
inghourswatchingvideoofhitters’preferences.Whenhis
playingdayswereover,hethoughtthathispitch-sequencing
system could be adapted to predictthesuccessofminor
leaguers. In 2015 he began tryingtosellinvestorsona fund
thatwouldpayminorleagueplayersinexchangefora share
oftheirpotentialfuturemajorleagueearnings.
A serendipitous meeting with PaulDePodesta—acharacterin
Moneyball in his former role as assistanttoOaklandA’sgeneral
manager Billy Beane, now part oftheClevelandBrownsfront
office—led to a sit-down with Miller.Notlongafter,Schwimer
had secured $26 million. In 2016hefoundedBigLeague
Advance and invested in 77 players.Hesaysthat 38 oftheorig-
inal signees are in the big leagues,witha dozenlikelytofollow.
In 2017, Schwimer raised an additional$130milliontoadd
to his roster, which now numbersabout 230 players.Hetook
themanagementfeesfromthatroundtoassemblewhathe
calls“theworld’sbestprivatedataanalyticsteam”insports,
andJambosAnalyticsLLCwasborn.Thenameis a bro-tastic
acronym of six original members.Therearenowabout15.
Jambos decided to sell picks inbundles: 17 weeksfor$3,000,
4 weeks for $900, and 1 week for$250,whichis whatI signed
up for. Each bundle came with apaybackguaranteeif thesys-
tem wasn’t profitable. It pays $10,000tofull-packagecustomers,
$2,000 to monthly customers, and$300toweeklycustomers.
And if Jambos was up, but not enoughtocoversubscription
costs, it would make up the difference,plus$100.
Day by day, I got better at usingtheservice.Still,because
of discrepancies in the lines andodds,I wasn’tgettingbetsin
on about a third of the games. Sowhatwasthepointinusing
Jambos? My results would havemoretodowithwhichbetsResultsofbetting 111 picksfromDec. 31 toJan. 6
Win Loss Push
NobetmadeTotalwagers
Netwinnings$2,250$63.60I placed than with how well the company’s models worked.
I wasn’t alone. On New Year’s Day, Schwimer tweeted, “We
will have to stop selling picks very soon.” Professional bet-
tors and books, he said, were building bots to react to Jambos
picks and moving lines before customers could act. “The
vast majority of our recommendations were moving very
quickly,” says Schwimer. The last bundle of Jambos picks will
end with the Super Bowl, after which it will continue to pub-
lish some online for free.
Schwimer expected this to happen, he says, but not as
quickly as it did. He was, in his telling, a victim of his own suc-
cess, and I saw anecdotal evidence to back his claim. On a cou-
ple of occasions, lines moved against me in the time between
entering my bet and tapping to confirm it. But more often than
not, I could get my wagers in, and I didn’t have bots working
for me. On 19 bets, despite my hunt-and-peck method, I got
better lines than Jambos recommended.
Schwimer had another reason for shutting down the sub-
scription service, which he describes as only a small part of
the larger company. There were a “handful of people,” he says,
who didn’t have the bankroll to continue betting after sustain-
ing heavy losses early. Most customers, he says, reported mak-
ing money. (He declined to disclose subscriber numbers or
detail company finances.) The most satisfied customers, he
says, were those who had a sizable bankroll and treated Jambos
like a financial product: “They understood that this is an invest-
ment. This is not like betting where it’s a fun thing.”
He was right about the fun. Around the time on a Saturday
morning that I was searching in vain for the Jacksonville State
vs. Eastern Kentucky basketball game at under 146 points, I
began to regret volunteering to be a betting degenerate. I was
mostly updating spreadsheets. It felt a lot like work.
In the end, I made money. Of the 111 bets Jambos recom-
mended, I made 75. Of those, 40 were winners and one was a
push—a 54% success rate. Three times, since I found more favor-
able lines, I won bets that Jambos pushed on. Of the 36 I didn’t
make, only 16 were winners for Jambos. This proved to be my
salvation. Jambos’s record for the week was 53 wins, 54 losses,
andfourpushes,disastrousformostcustomers.Inmycase,
bettingatjust$30a pop,therebatepaddedmy$63.60inwin-
nings,leavingme$113.60ahead.
Whatdotheseresultsindicate?Notmuch.Beforemyexper-
iment,I wasconfidentthatJambos’salgojockeysknewthebet-
tingmarketsbetterthanI did—anddoubtedtheyknewmore
thaneverybodyelse.Aftera week,I feltthesame.DoesJambos
havegoodalgorithms?Probably.Butsodoa lotofpeople.
Togethertheycreatea fairlyefficientmarket.Asforentertain-
mentvalue,nooneneedshelpfromJambos,oranybodyelse,to
blowmoneybettingonWinthropUniversitybasketballgames.
WiththedeathofJambosasa pick-selling service, the
tout industry remains a swamp of small-time operators. But
Schwimer says Jambos isn’t done yet. He has big plans com-
ing in the spring. “We’re about to make a huge splash in the
industry,” he says. “People won’t believe until they see it.” <BW>
�With Eben Novy-WilliamsGAMBLING WITH JAMBOSDATA: JAMBOS, BLOOMBERG REPORTING