Technical analysis is the examination of past price movements of shares in order to predict future value. This information should help the technical analyst (or chartist) buy and sell at the right times with as low a level of risk as possible. In adopting this method the analyst does not worry about what is being bought and sold or even what factors influence the price except purely market sentiment. Who cares if Yahoo! barely makes a profit while it is more capitalised than the Royal Navy - if there is an upwards momentum there is the chance to make money through speculation.

Technical analysis can be applied over long or short-term time frames, be based on current (lagging indicators) or only past (leading indicators) prices, on shares of large or small companies with upwards, downwards or cyclical price trajectories. Ideally it would be in strongly traded shares with a consistent pattern. Yet just as there is a huge variation in types of companies and market behaviour, a chart may illustrate many patterns appearing. The trick is to know which patterns are the most relevant and valid, while ignoring other price fluctuations as 'noise'.

Indicators

Some common tools of technical analysis include:

Moving Averages

Simple and commonly used, but by no means tells the entire picture. Moving averages look at how does the share price of company XYZ compare with its average price (usually closing price) over a timeframe (say 30 days, although day tradeers might look for the past 20 minutes). Some analysts may choose to use weighted or exponential moving averages to highlight more recent prices.
The shorter the timeframe, the more the moving average will mirror the actual price and therefore look more volatile than if it was calculated over a longer period. However longer periods may mask important trends.
A moving average can be used in three ways. Most simply, and for long-term trading decisions should not be based on daily fluctuations, shares are bought when prices trend upwards, and sold when they trend downwards.
A smarter use of a moving average is to compare it with the actual price of the share. When the actual price is higher than the moving average, it is time to buy in. Once the price stops growing and wavers around a price level, the rising moving average line will converge with the stagnant acutal price line, and that would be a good time to sell.
A third option would be to use two or more moving averages, with one average - the 'slow' average - being based on a very long time frame (say 22 weeks, or 110 days). The chartist would then buy when the 'fast' average exceeds the 'slow' average, and sell when the fast average dips below the slow average.

Bollinger Bands

Named after John Bollinger, the Bollinger Band is the space between the 'upper level' (say, two standard deviations above the simple moving average) and the 'lower level' (of equal proportion, i.e.: two below). Anybody with a faint knowledge of statistics know that anything beyond the second standard deviation of anything is pretty flukey (top/bottom 5%).
One use of Bollinger Bands is to detect sharp trending (which direction is another matter - you need other tools and research to answer that). Explosive growth or falls often comes after a period of quiet trading, which on a Bollinger chart appears where the upper and lower levels are relatively close. Bollinger Bands can also define targets - buy when a moving average is within the lower half of the band, and sell when it enters the higher half.

Moving Average Convergence Divergence (MACD)

The MACD is the difference between one exponential moving average (EMA) and another EMA covering a longer time period. When the MACD is positive (i.e.: recently the stock is trading more positively than before), the market can be considered bullish.