Whitewater roulette

The paddlers scout the waterfall, their kayaks as bright as bits of candy against the Sierra Nevada’s sculpted gray granite. Behind them, water blasts into the mountain air through a misshapen rock spout, then hammers down on the rock ledge below. But Seth Warren’s eyes are on Phil Boyer’s hands.

Boyer, 35, is a two-time whitewater rafting world champion. Warren’s home is a van he drives across the nation on behalf of a sports clothing company. They speak in the soft patois of surfers, earn less money than many fast-food workers, can finish a jug of cheap wine in less than 20 minutes and kayak nearly every day. Their lives are the envy of river aficionados worldwide.

Today they stand over “the teacups,” a largely vertical stretch of the Kern River, because, when it comes to running rapids and assessing risks, they are sure they’re among the nation’s best. They also know the river could kill them anyway. But that snippet of rational thought may well be lost at the moment, roiling as it does amid so much sensory input, hormone-fired emotion and more immediately significant information.

Approach from the left, then charge hard to the right, Boyer says, mapping out the run with hands that have become rock, water and kayak.

Warren, 26, watches and listens intently. To a point.

“Don’t overexplain the rapid to me,” he finally cautions. “I won’t be able to run it.”

Random fatalities

Warren and his companions — Boyer, Morten Eilertsen and Mariann Saether — know a kayaker broke both his arms here two years ago. Boaters die in similar waterfalls every year. In 2000, the last year for which complete statistics were collected by American Whitewater, a river-sports advocacy group, at least 47 paddlers drowned nationwide. Some perished because of inexperience or lack of safety equipment, but many of the deaths conform to no pattern.

In one case two kayakers successfully cleared a 6-foot waterfall and slid through a rock-infested landing zone, according to incident reports. The third kayaker, however, slightly misguided his boat and ran it into a rock. Water pressure pushed the craft against a boulder, flipping it over and plunging his head underwater. The impact was so great that it collapsed the top of his craft, clamping his leg and preventing escape.

For more than an hour the kayaker’s friends swam to the pinned boat, attaching a rope and trying to drag it out. The water’s pressure would not allow release of the craft. A moment after the survivors decided to give up, the current shifted slightly and the boat popped free, exposing their dead friend. It took four men using tools to remove the body from the kayak.

In another instance, a kayaker easily surmounted a waterfall in North Carolina amid rain, sleet and high winds. Fifty yards downstream, a small ledge in the riverbed created a “hole” — a depression in the river’s bottom where water dangerously recirculates, as if in a washing machine.

The six other kayakers accompanying the paddler approached the hole only seconds behind him and saw his helmet flash in the backwash. He had been pulled from his boat by the river’s force. The other boaters threw rescue bags with ropes attached, but the swimmer did not grab for them, apparently already unconscious.

He recirculated for four minutes, each cycle holding him under for 20 seconds, allowing him to surface briefly, then pulling him under again. Then he popped out of the hole and floated serenely, already dead. None of the others even brushed danger.

Each year almost as many people die on calm, moderately challenging rapids as on large, expert-only waterfalls. The river’s ever-changing flow creates and dismisses an infinity of potential hazards every second, professional kayakers say.

Warren and his companions believe they know how to identify and avoid risks. And to a degree, they are right — years of experience and physical finesse increase their safety. But much of their confidence is an illusion. While humans are exceptionally good at identifying risks — owing to our unique capacities of pattern recognition — we’re notoriously fallible at estimating the likelihood that a risk will materialize.

In kayaking, hazards lie beneath the water’s surface and risks are unpredictable. If one of these kayakers dies, it will likely have far less to do with his abilities than with the fact that we humans are, at our core, awful risk assessors.

Risk aversion

“I have a plan for every stroke before I even get wet,” Warren says as his bright orange boat glides through the first teacup, paddles flashing like the blades of a windmill.

As he and Boyer approach the drop — an 8-foot tube of water that angles at 45 degrees — they are confronted with a choice. To the left of the lip, water falls calmly but a small rock juts out, a slight though obvious hazard. To the right, the water churns viciously. There may be dangers underneath, but they are hidden from sight.

As they analyze the options, the human weaknesses that plague decision making emerge.

The mistake Warren is about to commit was discovered decades ago in a series of experiments. In one of the first social science laboratories, professors in white coats recruited human lab rats and presented them with two alternatives: a traditional insurance plan, in which the premium is $100 and all stolen belongings are replaced, or a new type of insurance in which subjects pay only a $45 premium, but if their belongings go missing the insurance company will flip a coin. If it comes up heads they’ll pay the claim, but tails pays nothing.

From a rational perspective, the $45 choice is superior: Subjects received 50% coverage for only 45% of the premium. But subjects overwhelmingly chose the $100 insurance policy. Why? Because most humans instinctively seek to eliminate risk rather than merely reduce it.

Warren’s natural instinct, like that of people choosing between insurance plans, is to eliminate the river’s risk. But this desire is so strong that the slightest of cues can encourage the perception that risk has been removed.

As Warren and Boyer paddle forward they begin subconsciously classifying obstacles as either risky or safe. Their natural instinct is to choose the path that seems completely riskless, rather than the avenue that contains small, potentially risky obstacles that can be avoided.

The greatest hazards in whitewater kayaking are unseen. A decade ago Boyer surmounted a waterfall when his boat suddenly stopped, the bow pinned beneath an underwater log. His boat slid farther under the log until the wood pushed against his body, forcing him underwater and preventing his escape. All he saw were white bubbles, then green light as his head dropped below the surface.

Suddenly a current caught the back of the boat and pulled it free. If not for the chance surge, he would have died. From the surface, the log was invisible.

On the teacup, the turbulent water to the right might hide anything. But the instinct to eliminate risk, rather than tolerate a reduced hazard, pushes them toward the frothing portion of the fall.

“I couldn’t tell what was going on to the right, but I assumed it was safer,” Boyer says later.

Both kayaks twist slightly as they enter the waterfall, the boats rocking as the underwater hydraulics push and pull. Then they shoot into the air and plop into the next pool. There have been no underwater obstacles. After a brief delay, both men successfully surmount the second waterfall, riding the water as it sprays over a rock protruding like a giant’s brow, and pause in the cold water beneath.

Overconfidence

Warren and Boyer paddle to the lip of the next waterfall and look at each other.

“Do you remember looking at this drop?” Boyer asks.

“Uhh, no,” Warren replies.

If they could step onto the rocky bank, they would see an 11-foot fall, the rushing water filtering the light like sea glass and making the rocks underneath pale and wavy. But from their kayaks, all they can see is a thin line where the drop begins, and the horizon. They stood above this fall only 10 minutes ago, trying to memorize all six drops. Now their memories are blank.

“Well, I guess that means it’s OK,” Boyer says hopefully, and he paddles to the edge.

Certainty is an integral part of kayaking. “You have to be 100% confident,” Boyer explained earlier. Every kayaker feels a little bit of fear on the water — that’s part of the excitement. “But if you start thinking about the mistakes you can make, you’ll make them.”

But confidence can be dangerous. After the white-coated experimenters learned about humans’ instinct to eliminate risk, they began to ask why. One answer is overconfidence — a compensatory instinct that pops up when humans crave control but are confronted by random chance.

Boyer’s confidence in his abilities may be so high because he has so little control over the river’s dangers. In a series of experiments, professors tested how people respond to random events: Subjects were shown reports on 12 stocks and asked to predict the stock market’s future, and to rate their confidence in their predictions. Researchers found that if the average stock buyer is asked to predict, say, the price of Wal-Mart’s stock over one year, most will admit ignorance. But when confronted with a new high-flying Internet offering and lots of meaningless tables and charts, many will make predictions with unwarranted confidence.

For his part, Boyer seems as confident as the granite as he teeters atop the fall and smoothly drops over.

“No problem!” he shouts up to Warren, who follows him over the edge.

The big fall

As they approach the big waterfall, Warren doesn’t think about the warnings Boyer mentioned earlier. Instead he focuses on not focusing and dips his paddle deep into the rushing water.

Psychology supports Warren’s instinct. Eventually researchers began wondering: If everyone is prone to overconfidence, why do some people consistently make the right call?

The answer, they discovered, may have to do with which part of Warren’s brain is doing the heavy lifting. The prime goal of decision making is to move thought from the prefrontal lobes of the cerebral cortex, where deliberate, conscious decisions are made, to the posterior neocortex and amygdala, where thought processes are more automatic. Athletes refer to it as being “in the zone”; psychologists call it “flow.”

As he approaches the big fall, Warren scans the water almost lazily, paddling into the waterfall’s spray and twisting his body as the kayak goes over the drop. He lands softly. “I tried not to think too much,” he shouts up as explanation.

Warren and Boyer know they paddle better when they are in “the zone.” But as their thinking becomes less deliberate, it falls prey to a psychological weakness: the subconscious perception of patterns where none exist. They believe that previous successful runs prove they are ready for these dangerous rapids. But the truth is that their previous successes may inspire them to believe in patterns that blind them to dangers that can quickly emerge.

The teacup falls are similar to rows of horizontal and vertical lights laboratory geeks used in the 1950s to discover what happens when different parts of the brain try to predict the future. Subjects were shown the lights and asked to predict which row would light up. They were told the lightings would be random.

In spite of this knowledge, subjects of the study were incapable of choosing randomly. Rather, they were influenced by whether their previous prediction had been correct. If so, there was up to a 72% chance that they would make the same guess. Even when subjects were offered financial rewards if they chose randomly, they were unable to do it. Subconsciously, subjects saw patterns where none existed and were influenced by a cause-and-effect relationship that wasn’t real. The same instinct enriches casinos every day.

As they think about kayaking, Warren and Boyer perceive patterns. “I’ve run a thousand falls,” Warren says, “and they’ve gotten me ready for these ones.”

Warren is right. Years of practice have prepared him to confront bigger and more deadly risks — as long as those risks increase incrementally. But if the risks suddenly become exponentially more deadly, his instinct to rely on patterns may expose him to danger, says Baruch Fischoff, a professor of social and decision sciences at Carnegie Mellon University in Pittsburgh.

As a river rises from 1,500 cubic feet per second (cfs) to 2,000 cfs, it may become marginally more risky. But the same river may be significantly more deadly as soon as it hits 2,500 cfs. To someone seeking out patterns, it appears the flow has only increased incrementally. Skeptics know that the risk has surmounted a tipping point and is now exponentially more dangerous.

Boyer and Warren both have an instinct to base future predictions on patterns, and they are both experienced kayakers. But they have never run these waterfalls before.

“People typically imagine they are more in control then they really are,” says Fischoff.

Boyer concentrates as he approaches the big fall. Others have warned him to push his boat high onto the rock wall and use it to slide over the turbulent flow. He focuses on the granite walls, hardly looking at the water around him, and paddles toward the incline.

His boat tips over the fall, rides up the rock and stalls, the bow almost out of the water. The river whips the stern around. Boyer twists hard to compensate, but by then the kayak is jutting over the drop, teetering. It pushes over the edge and falls, rotating clockwise through the air as it builds speed, then smacks into a ledge, bouncing and beginning to flip. Boyer loses control. The kayak corkscrews. His head barely missing the rock, he plummets over the falls upside down and lands that way in the water.

A vicious twist propels his body upright.

“I missed it!” he screams.

He runs his hand up his neck, making sure he is uninjured.

Moments later, the men pause in the chain’s final pool.

“Should we do it again?” Boyer asks as they pull their kayaks onto stone.

“Sure,” Warren says. He looks up, scanning the immutable rock and ever-changing water. “These falls are a no-brainer,” he says. The men start climbing.

---

Charles Duhigg can be reached at charles.duhigg@latimes.com.