This is a history of the financial quant world. The time line extends from financial trading as it existed in the 1980s (i.e. human traders shouting at one another in the pit, writing down orders, getting buyers and sellers together through verbal means, etc) to the pervasive high-speed trading that occurs today.

The old guard was out maneuvered as new technology came into the frame; being mostly in denial about the new comers while claiming that stock trading was a game that required mono-a-mano interaction. Eventually, of course, the old guard had to change or be shown to the door.

As an engineer, I enjoyed reading how the first pre-networking programmers crudely automated the trading system via personal computers linked by connections. 

I never really understood how you could make money at high-speed trading. The whole concept seems to take advantage of idiosyncrasies of the particular plumbing that makes up the market. For example, one exchange is connected to another via a pipe that isn't quite as fast as another so, if you see movement on a stock on one exchange, you can buy a stock on one exchange then sell it quickly on the other exchange for a different price, pocketing the difference. There are other techniques which involve more adversarial actions.

Orders are supposed to be executed in a first-in, first-serviced queue. However, traders uncovered bugs in the system that allowed them to jump to the beginning of the queue and get their orders placed first.

Mr. Patterson also discusses how the people running market have adjusted, segregating the stock market into “lit” and “dark”pools. The purpose of the ''dark pools” is mostly to hide transactions from the open market. If particular transactions show up on the open market, the high-frequency traders will take note and make trades that will spoil the profits of the original trader.

So, there's a lot of money made here but where does it come from? There are many strategies but an early modus operandi was to watch for someone (perhaps a mutual fund) looking to buy an asset. The bots sense the impending deal and buy before the fund (at the current price). When the mutual fund buys the asset, the bot offers the stock to the fund at a slightly higher price than its purchase price. The mutual fund thus pays a higher price that it would have without the bot being in play. The mutual fund loses.

This is a good, understandable book on a complicated subject. Mr. Patterson makes it readable and enjoyable.