Author Archives: Thomas Carlsson

Rio 2016: Men’s 1500m final

This was a race worth analysing.

Race Video

Centrowitz’s victory in this race was obviously very unusual in that he led (almost) from start to finish. Clearly there wasn’t much tactics involved in what he did  – he just kept everyone at bay and countered every attempted overtaking by increasing his speed. This obviously attests that he was the fastest runner in the field on this day. But even if that more or less explains why he won, it’s still interesting to take a look at what Centrowitz’s opponents, especially Kiprop, Makhloufi and Souleiman, did during the race. There are a few clear reasons why they could not do better in the sprint.

There is essentially no movement at all in the field of athletes for the first 600m of the jog stage. The pace is very slow and the race is effectively reduced to an 800m competition.

After jogging at the back of the group in the beginning, Kiprop moves to the front on the 600-700 straight. This is an excellent tactical move because the pace still remains slow and he easily overtakes the entire group. Waiting to the 800-900 straight before making his move would have done Kiprop no good at all.

As he moves to the front next to Centrowitz, Kiprop seems to stumble a little, which allows Blankenship to overtake him and leaves Kiprop in a boxed position, but he is still well-positioned for the sprint.

Kiprop’s move drops Makhloufi almost to last place. As the pace remains very slow on the 800-900 straight, both Kiprop and Makhloufi run passively and let other opponents improve their position. At the 900m mark Kiprop is boxed into 9th position while Makhloufi is 11th. The task ahead of them grows every second, because every runner in the field is still in good strengths and the final lap is obviously going to be very fast. At this point Kiprop and Makhloufi have squandered their opportunities for overtaking opponents at slow speeds, and will instead have to do a lot of overtaking at speeds close to 13 seconds / 100m to win this race.

It should be remembered that Makhloufi had already run three rounds in the 800m and two rounds in the 1500m prior to this race, so he was probably not feeling very energetic in this race. Saving as much as possible for the sprint made some sense, but the pace set by Centrowitz turned out to be so extremely slow that the sprint became very crowded in the middle of the group.

On the 1000-1100 straight Makhloufi wisely moves to the outside, but then foolishly fails to move all the way to the front at once. After a 2.45 split at 1000m, it should be pretty clear that the sprint is about to start any second. Kiprop comes surging from the back of the group and tussles with Makhloufi for the front position, costing them both some energy.

Meanwhile, Souleiman has run a smart race up to this point. After following Kiprop’s move on the 600-700 straight, he smartly holds his position and moves up on Centrowitz’s shoulder when he is threatened from behind at the 900m mark. At the 1000m mark he takes the lead for a brief moment. But he, too, runs a bit thoughtlessly on the 1000-1100 straight as he lets Centrowitz pass him on the inside, instead of closing the gap. Maybe he feels that it’s still too early for him to really assert himself and go for the victory. But just a few second later, the sprint begins as Makhloufi and Kiprop come charging on the outside, and Souleiman is immediately boxed, whereas Centrowitz escapes the box. After a slow start, the race combusts in a split second.

All three of these athletes (Kiprop, Makhloufi and Souleiman) still had a chance to outrun Centrowitz in the final lap, but none could. As I said, Centrowitz’s victory was more due to his better running ability than to smart tactics. His three main opponents did make a few tactical mistakes which cost them some energy, but this was still anybody’s race with 100m to go.

 

Rio 2016 Women’s 1500m heats: overtake when it’s slow

Race videos

I will add videos here if I ever find them.

Analyses

These races were fascinating from a tactical perspective. Races with 14 participants are crowded and it was interesting to see how runners dealt with this challenge.

Heat 1 starts so slowly that Nancy Chepkwemoi decides to overtake the entire field in the first bend. That’s basically a sound decision, but look what happens after that. The pace remains slow, and it seems like almost every other runner in the field makes the same decision as Chepkwemoi did. There’s constant stream of athletes moving towards the front in the outside lanes. I haven’t seen a race with this much movement through the whole field for a long time.

Most of these overtakings look like good decisions. If you can pass the entire field easily, do it! With everyone moving up on the outside, Chepkewmoi falls back almost to last place. The reason for this seems to be that she runs tentatively and constantly leaves a gap between herself and the runner in front of her, and others keep moving into that gap. Chepkwemoi then has to make a panic move to the outside at the 1000m mark, runs the 1100-1200 in lane 2 and then runs out of energy.

Dibaba displays great tactics in heat 1, moving to second place at 700m and staying there until the sprint. Perhaps her reputation helps her a little bit, since the runners behind her clearly prefer to stay there, and do not try to move ahead of her even when the pace remains relatively slow through the 700-1100 lap.

Brenda Martinez runs the whole race out in lane 2. Even as many other athletes move past her towards the front, she always stays put in the middle of the field. I suppose she wants to make sure she’s in a free position when the sprint begins, but she clearly runs the longest distance of anyone in this race and does not look very comfortable sprinting on the final straight. It is often preferable to run at least some parts of the race in lane 1, either at the front or at the back as Dibaba did. In the semifinals Martinez will have to save more energy for the sprint if she wants to make her way to the final.

Heats 2 and 3 were both fast and therefore non-tactical. The pacemaker from heat 2 did not qualify, but Sifuentes, who set the pace in heat 3, did go through to the semifinals.

Rio 2016 Men’s 8oom heats: follow the favorite!

Time to take a quick look at the middle-distance races in Rio with an analytical eye.

Race videos

I will add videos here if I ever find them.

Analyses

Fast heats have become very typical in the men’s 800m heats in this decade. More often than not, the runner who sets a fast pace is one of the favorites in the race. Other runners should therefore have a very simple goal in their minds: just follow the favorite.

Heat 3 in the men’s 800m provides a great example of one athlete who finds success by following the favorite, and another who fails to follow the favorite and runs a miserable race. van Rensburg from South Africa is the first athlete and Murphy from the USA is the second.

Presumably these athletes have seen David Rudisha run on TV before facing him in this heat. With that viewing experience in mind, one would think that their pre-race strategy for the 100-200 straight should be to compete for second place in lane 1, behind his back.

It’s interesting to watch van Rensburg and Muprhy as they come out of the first curve. They’re essentially even and they can already see that Rudisha will take the lead. Murphy does not recognize the tactical importance of this moment, so he let’s van Rensburg go ahead and take second place. That’s not a big mistake in itself, but he then moves out to lane 2, gets into all kinds of trouble and wastes a great amount of energy just to keep up with the pace.

van Rensburg, on the other hand, runs the perfect race for an underdog with a 1.46 season best. He just stays behind Rudisha all the way and saves his strength. He gets a bit lucky on the 500-600 straight when Rudisha accelerates and leaves van Rensburg a few steps behind. Someone might fill the gap between them…but the athlete next to van Rensburg happens to be Murphy, who certainly doesn’t have any energy to spare now, after his adventures in the outside lanes. So van Rensburg gets a free path in the inside lane all the way and secures qualification, while Murphy struggles to fourth place but still qualifies as a fast loser.

It will be interesting to see if Murphy learns his lesson: follow the favorite and fight for second place behind him on the 100-200 straight.

Heat 4 provides another good example of how to follow the leader. Alfred Kipketer may not be Rudisha, but he has run sub 1.44 this year, more than a second faster than anyone else in this heat. He goes out fast and Algeria’s Hethat takes second place behind him. At the 200m mark Hethat has clearly set himself up for an easy qualification: he has the second fastest season best in this field and he’s following the favorite. What could go wrong?

But then Hethat loses his nerve at the 300m mark, presumably because he finds the pace too slow. He intentionally leaves a gap to Kipketer and pushes his way outward. But Kipketer is still keeping a good pace at 53,5 seconds, so it’s clearly a bad decision for Hethat to give up his position behind the favorite. The Danish runner Bube recognizes the open gap ahead of him and moves past Hethat in lane 1, into second place behind Kipketer.

Bube qualifies with relative ease even though he only had the 7th fastest season best in this heat. Hethat performs a great 300m sprint in the second lane, but only secures his qualification with a very narrow margin and maximum effort.

Heat 5 provides an interesting example of what might happen when the early leader is not a favorite. The Spaniard de Arriba, who leads through the first lap, is overtaken by every opponent on the 500-600 back straight. Fortunately for the others, they all show good tactical awareness and seem to recognize what is about to happen on the back straight. Everyone moves outward to avoid getting stuck behind de Arriba.

 

A first introduction to patent criteria

Patent law rests on two fundamental assumptions:

  1. Technical progress is a good thing and should be promoted.
  2. Most inventors will keep their inventions to themselves unless given a strong monetary incentive to make them public.

Neither of these assumptions is unchallengeable. Progress in weapons and gene technology, for instance, can be morally questionable, while some inventors, such as university researchers and freelance software developers, would disseminate their inventions even without any additional incentive. Nevertheless, assumption 1 holds true in most areas of technology and assumption 2 holds true for most inventions made in commercial industry. The assumptions are thereby presumably true enough to at least justify the existence of a patent system.

This justification can be stated as follows: a patent is an exchange between a private inventor and the general public, beneficial to both parties. The inventor agrees to publish his invention in such detail that the patent publication (where the invention is described) allows other inventors to learn from it and improve upon it. This accelerates technical development because publication reduces overlapping and duplicate inventive work. In return the inventor may be granted the right (for a limited time, in a limited geographical area and for gradually increasing fees) to prohibit others from selling, manufacturing or using the invention.

The inventor faces many uncertainties with respect to his private monetary benefits. Quite often the invention becomes public but the inventor receives only a rejection letter in return. If patent applications were to become irrevocably public immediately upon filing, this fundamental uncertainty would probably be a very strong deterrent against filing anything in the first place. To encourage active filing, patent applications do not become public until 18 months after their filing date. The application is examined at least once in this period, and the inventor is allowed to withdraw his application before it becomes public. The inventor may consider withdrawal especially if the first examination turns up strong objections against patentability. The possibility of withdrawal reduces uncertainty a lot, but not to zero. Sometimes the inventor is given an initial green light in the first examination, but damning objections are discovered after the publication date. By then it is too late to retract the application.

The benefits accruing to the inventor in the publication-for-exclusive-right exchange are therefore not automatic or certain. The patent may not be granted at all, it may be granted only for a very restricted part of the invention, or a green light before the invention becomes public may turn into red light after it has been irrevocably published. However, the public benefits of the patent exchange are not automatic, either. In fact the reason why the private benefits are so uncertain is precisely that complicated criteria have been put in place to protect the public benefits. Applications must meet these criteria if a patent is to be granted and a lot of examination and argument is therefore often required before a final decision can be made. There are three substantive patent criteria that are centrally important: inventive stepclarity and sufficient disclosure.

Inventive step:

Like any other exclusive right, every granted patent prevents competition (albeit only in the very limited subfield of technology defined by the patent claims). It is therefore clear that not every application can be granted. Inventive and industrial activity would grind to a halt if every field of technology would be laden with monopolies. It would also place a great burden on the courts where patent disputes are resolved. This is why applications have to be examined by patent offices, and some of them have to be rejected.

On the other hand, it should not be overly difficult to obtain a patent because there must be a sufficient incentive for inventors to disclose their inventions to the public. Patent offices have to separate useful and unique inventions from ones which merely repeat technical solutions which have been discovered before. To do this in a rational and informed manner, they judge every claimed invention by so-called inventive step criteria.

In very simplified form, the criterion for inventive step is that the same invention (or a very similar one) should not have been published before anywhere in the world. A direct copy of someone else’s invention can therefore not be patented (at least in principle – it is of course in practice very difficult to compare prior publications comprehensively with a global scope). But more importantly, inventive step criteria separate inventions which present only small, trivial or obvious improvements from the ones whose inventive contribution is large, significant, or surprising. The expression “very similar” in the parenthesis above is therefore at the heart of the discussion whenever inventive step is disputed.

Clarity and sufficient disclosure:

The public benefits of the patent system also suffer because some people abuse the patent system.The forms of abuse are multifarious, but one is particularly important. Applicants may present a so-called invention in deliberately obfuscated form. An invention defined in slippery terminology is problematic from a public-benefit perspective because it can be interpreted in many ways. The application process becomes muddled because comparisons to previous publications can be difficult to make. If the application happens to be granted with an ambiguous scope, the door is opened to creatively argued infringement lawsuits. In order to prevent such abuses of the patent system, applications have to meet the two remaining criteria: clarity and sufficient disclosure.

The patent claims are the centerpiece of a patent publication and they are always the primary focus of the clarity assessments performed by patent offices. The criterion of sufficient disclosure is more concerned with other parts of the patent application, especially the description and the figures. Patent publications will not be useful to the public (and will probably not even be read) unless they reveal a sufficient number of useful details about the invention. The quality control exercised by patent offices on sufficient disclosure is sometimes only cursory, because the primary reasons for rejection or acceptance nearly always relate to the patent claims, to inventive step and clarity. But in subsequent essays I will argue that all three of these criteria are in fact closely intertwined and cannot be separated from each other.The sufficient disclosure criterion is fundamentally important for realizing and balancing the private and public benefits of the patent system.

 

Beijing 2015 Women 800 and Men 1500m finals

Finally, here are tactical analyses of last weekend’s finals.

Race videos

Women 800m

Men 1500m

Analyses

Women 800m

Based on the heats I expected to this to be a very fast and non-tactical race, but it was actually slower than expected. Bishop probably makes a tactical error on the 300-400 straight when she lets Lamote overtake her without any resistance. It is  an understandable error. Bishop was probably expecting Sum to start very fast, and indeed she did run the first 200m in just over 27 seconds. But Sum then slowed down the pace and the second 200m took 32 seconds. Bishops fails to read this pace change and stays in the inside lane. She should probably have moved up next to Arzamasova to discourage Lamote from going past. This would have allowed her to stay closer to Arzamasova in the sprint.

Positions remain unchanged all the way to the 600m mark. Bishop remains in a partly boxed position until final sprint and seems to be losing out, but then Arzamasova suddenly gives her a free route on the inside, a huge tactical error by the eventual world champion. Bishop comes close to winning the race but is unable to capitalize on this mistake.

Men 1500m

As I expected the other Kenyans take the lead in the start while Kiprop stays at the back. The pace isn’t fast by any means but everyone remains content with their positions up to the 800m mark when Centrowitz finds a way out of the red queue and moves into green position. Other athletes follow him on the outside but the pace remains slow to 1000m.

Most athletes probably expected Kiprop to move towards the front at this point, but he remains at the back of the group. Iguider and Makhloufi both move up on the outside while Manangoi successfully defends his lead position against Centrowitz’s challenge. At 1100m Centrowitz looks set for a medal. Iguider and Makhloufi are also in great position, although they have ran much of the race out in lane 2. Kiprop is in bad position behind the entire group and has left himself a lot of work to do in the final lap.

Centrowitz begins to fade on the back straight and inhibits other the athletes behind him from sprinting effectively, although Manangoi and Kiplagat both manage to escape on the inside before Centrowitz blocks it. Makhloufi leads the sprint from the green queue while Kiprop comes storming from the back. The final straight is a classic battle where no tactical moves are needed.

Most of the finalists seemed to be worn out from the heats and semifinals since they ran out of energy pretty quickly in the sprint. This made it fairly easy for Kiprop to overtake them all on the 1200-1300 straight. Kiprop was so much better than everyone else that he had much room for error and probably could have won this race with several different strategies.