Aims. We aim to give a reliable estimate of the number of flaring giant stars in the Kepler field. By analyzing the flaring activity of these stars, we explored their flare statistics and the released flare energies. The role of oscillation in suppressing magnetic activity was also investigated. We searched for flaring specialities that may be associated with the giant nature across a sample of flaring giant stars. Methods. We searched for flares using the ≈4 yr long Kepler data on a sample of 706 stars compiled from two lists of flaring giants (log g ≤ 3.5) found in the literature. To lessen the probability of false positives two different pipelines are used independently for flare detection. Tests are carried out to correct the detection bias at low flare energies for a subsample of 19 further studied, frequently flaring stars. For these 19 stars flare energy distributions and flare frequency diagrams (FFDs) are constructed. For comparison purposes KIC 2852961 is re-analyzed with our present approach. Results. From the 706 Kepler flaring giant candidates, we ruled out those where oscillations or pulsations were misclassified and those that turned out to be dwarf stars. Finally, we confirm only 61 stars as flaring giants. Among these 61 flaring giants, we found only six that also show oscillations; we suggest that a large fraction of the 61 flaring giants are members of spectroscopic binaries, which has already been proven for 11 of them. The number of detected flares on giant stars correlate only weakly with the rotational periods. The FFDs for the 19 most flaring stars were fit by power-law functions. Regarding log–log representation, the slopes of the individual fits lead to an average α = 2.01 ± 0.16 power-law index, but the ranges of flare energies scatter within almost two orders, showing the inherent heterogeneity of the sample of flaring giants. Broken power-law fits are applied for two giant stars that have similar flare energy ranges; however, the energy at the breakpoints of the power laws are different, unveiling possible differences in the magnetic field strengths and atmospheric structures of these stars. The average power-law index of α ≈ 2 is the same for the flaring giants, the (super)flaring G-dwarfs, and dwarf stars between spectral types M6 and L0. Conclusions. The 61 confirmed flaring giant stars make up only ≈0.3% of the entire giant star population in the Kepler database, which is in contrast with previous estimates of about an order higher percentage. We found that most of the false positives are in fact oscillating red giants. No strong correlation was found between the stellar properties and the flaring characteristics. The majority of the flaring specialities are hardly related to the giant nature, if at all. This, together with the finding that the observed flare durations correlate with flare energies, regardless of the flare energy level and stellar luminosity class, suggests common background physics in flaring stars, or in other words, a general scaling effect behind the flares on different stars.