By Keith Idec

The Gennady Golovkin-Kell Brook fight Saturday turned out to be more competitive and entertaining than many skeptics expected.

Those that did tune in to watch it live – an average of 843,000 viewers, according to Nielsen Media Research – were treated to a back-and-forth battle Golovkin won by fifth-round technical knockout at O2 Arena in London. The live rating was lower than most Golovkin bouts because it aired live at earlier times than usual in the United States (about 6 p.m. ET and 3 p.m. PT).

The rebroadcast, however, drew an additional average of 593,000 viewers following HBO’s live coverage of another highly entertaining encounter between super flyweight star Roman “Chocolatito” Gonzalez and Mexico’s Carlos Cuadras. The re-broadcast rating doesn’t differentiate between new viewers and repeat viewers, so it is difficult to determine how close the combined number (1.43 million viewers) is to ratings Golovkin usually produces for his live HBO fights in prime time.

Gonzalez (46-0, 38 KOs) earned a hard-fought, unanimous-decision victory over Cuadras (35-1-1, 28 KOs) by scores of 117-111, 116-112 and 115-113 at The Forum in Inglewood, California. Nicaragua’s Gonzalez also won the WBC 115-pound championship in a 12-rounder that amounted to one of the best action fights of 2016. An average of 833,000 viewers watched the Gonzalez-Cuadras slugfest.

Kazakhstan’s Golovkin (36-0, 33 KOs) survived some competitive moments against the previously unbeaten Brook (36-1, 25 KOs) to stop the British star in the fifth round. Brook’s trainer, Brendan Ingle, stepped in to stop the bout because Brook was taking punishment from the hard-hitting Golovkin and wasn’t throwing back powerful punches.

It was later reported by multiple media outlets that Brook suffered a fractured eye socket against Golovkin, who retained his IBF, WBC and WBA “super” middleweight championships.

Keith Idec covers boxing for The Record and Herald News, of Woodland Park, N.J., and BoxingScene.com. He can be reached on Twitter @Idecboxing.